IL121201A

Cellular communications systems

Abstract

This record has no abstract on file.

Term

No projected expiry on record.

6 claims: 2 independent, 4 dependent

1
121201/2 CLAIMS What is claimed is:1. A method for increasing transmission gain to a mobile station of a mobile communications system, the method comprising: substantially simultaneously transmitting from two transmit antennas so as toform a radiation pattern that is characterized by a plurality of radiation lobes, each lobe beingcharacterized by a width inversely proportional to a distance between said antennas, an amplitudeof said lobes being bound by said radiation pattern of said antennas;determining a transmission direction to a mobile station;and aiming said pattern so as to produce a maximum in said transmission direction, thereby increasing transmission gain to said mobile station, and reducing scattering into saidmobile station from foreign objects.
4
Apparatus for increasing transmission gain to a mobile station of a mobilecommunications system, said apparatus comprising:two transmit antennas positioned together with a pair of diversity receiveantennas, wherein said two transmit antennas transmit substantially simultaneously so as to forma radiation pattern that is characterized by a plurality of radiation lobes, each lobe beingcharacterized by a width inversely proportional to a distance between said transmit antennas, anamplitude of said lobes being bound by said radiation pattern of said transmit antennas;and aiming apparatus that aims said pattern so as to produce a maximum in atransmission direction to a mobile station, thereby increasing transmission gain to said mobilestation, and reducing scattering into said mobile station from foreign objects. 24 121201/2

Independent claims2

672 paragraphs in 33 sections, as filed

jininbv mwpn nmvo

CELLULAR COMMUNICATIONS SYSTEMS

Joseph Shapira 027896 27896sha.doc 30-Jun-97

CELLULAR COMMUNICATIONS SYSTEMSFIELD OF THE INVENTION

The present invention relates to cellular wireless communications systemsgenerally and more particularly to apparatus and methods for cellular communications with basestations.

BACKGROUND OF THE INVENTION

Cellular multiple access communications date back to the early eighties. Thenineties witnessed an outburst of this type of service throughout the world and the introduction ofdigital technologies. The market is expected to soar and expand into Personal CommunicationServices (PCS), offering personal service, a host of value added features, and total personalmobility, indoors and outdoors. Broadband services are expected to emerge at the beginning ofthe next century. These may require a partial renewal of the network infrastructure.

Cellular mobile communication attempts to provide mobility, multi-user capacity(many independent users access the system), coverage (service is offered over a large contiguousarea) and grade and quality of service.

Cellular communications are generally limited by local codes to a range offrequencies. A widely used technique of cellular communications employs spatial isolation inorder to be able to reuse the same frequencies beyond a given range called a guard zone. Thecommunications of each user is maintained with a base station, whose antenna is elevated abovethe scenery in order to achieve a well defined and controlled coverage area. Sectorization isachieved by directive antennas that illuminate only one sector, thereby reducing interference,enhancing performance and reducing a pattern of frequency reuse.

Each sector of cellular communications is characterized by a number of calls perunit area, also called area capacity. Area capacity may be increased by reducing the cell size.Small cells that are positioned below roof tops in urban areas are called microcells. These uselower and smaller antennas. The cell hardware is more compact, and in some cases has lesscircuits. Another technique for microcells involves the antenna and RF circuitry only, remote fromthe cell equipment and connected via RF, fiber or microwave link, to the cell. Such anarrangement is especially attractive for operators in possession of RF or fiber trunking, likeCATV companies. A future trend to increase the capacity of large cells involves smart antennas.These are multibeam array antennas at the base stations, controlled to form narrow beams that arematched to the disposition of the desired user and the sources of interference. These are expected l to enhance the coverage and the capacity. The complexity involved in this technology is expected to be relieved with new cell architectures, including, among others, active antenna modules.

The network infrastructure of a typical modern cellular communications systemincludes a number of base stations, the actual number being related to the capacity required(measured in Erlangs, which is the number of fully occupied circuits), and to the coverage area.The base stations generally constitute about 80% of the network cost. A typical cost for a fullcapacity large cell base station is $500,000 - $1,000,000. The infrastructure also includesinterconnect trunking, which depends mainly on the total length of interconnect lines, andswitching fabric, which depends on the number of cells and calling load (measured in BHCA -Busy Hour Call Attempts). The cost of the basic service of providing airtime depends mainly onthe number of base stations and on their cost.

One of the problems of cellular communications systems is transmission losses.The transmit chain of a first generation base station consists of single carrier HPA’s, filtered,combined and relayed by a high power cable to the mast. The losses involved in the chain amountto 8 - 10 dB. The carrier spacing is restricted by the combiner to at least 600 KHz.

In an effort to cut down losses, second generation base stations were developed. Asecond generation base station includes a MCLPA - Multi Carrier Linear Power Amplifier. Thisreduces the losses and adds flexibility to the design of the carriers (frequency allocations). A lownoise amplifier (LNA) is used in the receive chain in the base station. The LNA reduces cablelosses which degrade the system noise figure. An additional receive antenna is typically used fordiversity. Recent installations place the LNA on the mast.

However, the MCLPA is an expensive part, running from $10,000 for a minicell toover $100,000 for a full capacity cell. Furthermore, MCLPA’s are currently supplied to the wholemarket by a limited number of vendors. The MCLPA’s from these vendors are available only in apower range of about 25 to 500 W.

SUMMARY OF THE INVENTION

The present invention seeks to provide a novel base station for cellular wirelesscommunications based on a modular structure.

The present invention includes an active radiator module (ARM) that serves as a basic transmit/receive module in a variety of cellular base station configurations. The active radiator modules follow the trend of cellular architecture development and are designed to meet both current and future needs. It is a novel approach that may reduce the cost of the base station while providing desired flexibility.

In the active radiator module system, a combined signal is transmitted in lowpower through a cable to a mast, where it redistributes to the active radiator modules. Thenumber of active radiator modules needed is a function of both the total effective radiated power(ERP) and gain required. The receive chain includes an LNA in each element, which reduces thenoise figure of the system. The same active radiator module can serve in microcells that requiresmall power and low gain antennas. A remote RF unit is the least expensive solution for microcells. Its applicability islimited by the cost of RF trunking. It is the preferred solution for operators that have an access tothe CATV or to fiber trunking already laid. This unit includes an amplifier, an LNA, and atransformer to the trunking band. This same module may be a part of a microcell or a picocell, butthe RF is included inside the package, while the antenna is typically separate. The modularstructure of the base station of the present invention provides readily upgradable base stationperformance at relatively low cost.

By way of example only, the present invention is described herein for certaincommonly-used frequency ranges, such as for cellular telephones or PCS. However, it isappreciated that the present invention is not limited to these frequency ranges and may be appliedto any set of frequencies.

There is thus provided in accordance with a preferred embodiment of the presentinvention, a modular cellular wireless communication base station including a plurality of activeradiator modules located at a desired antenna location, each module including at least one antennafor transmitting and receiving, a transmitter including a power amplifier, and a receiver, a beamforming network controlling the relative amplitudes and phases of each of the modules, and an RFfront end transmitting over a low power link with the plurality of active radiator modules via thebeam forming network and receiving over a lower power link via a low noise amplifier.

In accordance with a preferred embodiment of the present invention, the RF frontend is located remote from the plurality of modules. Preferably each module is self-enclosed.

Additionally in accordance with a preferred embodiment of the present invention atleast one of the active radiator modules comprises two separate transmit and receive antennaelements. Preferably the transmit and receive antenna elements are isolated from each other byabout 15-30 dB, most preferably by about 20 dB.

Further in accordance with a preferred embodiment of the present invention thebeam forming network is located adjacent the plurality of active radiator modules, one fortransmit and one for receive. 1

Still further in accordance with a preferred embodiment of the present invention,the modular cellular wireless communication base station includes a CATV up/down convertermodule. Preferably the CATV up/down converter module comprises a coaxial cable connected toa CATV network, the cable carrying a CATV forward link and reverse link. A CATV diplexer ispreferably provided that separates transmit and receive signals. The converter module preferablycomprises a mixer, a phased locked oscillator and a band pass filter, thereby to eliminate imageand low frequencies.

In accordance with a preferred embodiment of the present invention the RF frontend communicates with the beam forming network via a fiber optic link. In one embodiment, atleast two separate fibers separately carry transmitter and receiver signals. Alternatively, one fibercarries both transmitter and receiver signals, and a splitter and a filter are provided to split andfilter the signals.

Additionally in accordance with a preferred embodiment of the present invention,the transmitter amplifier comprises a first stage comprising a monolithic silicon gain stage and asecond stage comprising a hybrid packaged power amplifier.

Further in accordance with a preferred embodiment of the present invention atransmitter filter is provided that reduces transmitter wide band noise in a receiver band.Additionally or alternatively, a transmitter filter reduces spurious signals that interfere with areceiver channel of a cell.

Still further in accordance with a preferred embodiment of the present invention,there are provided a receiver amplifier and a receiver filter, wherein the receiver filter reduces atransmitter signal to a level wherein interfering intermod products are not generated in the receivechain, and the receiver amplifier is not desensitized by saturation. The other purpose of thereceiver filter is to reduce interfering signals from other base stations and other systems.

Yet further in accordance with a preferred embodiment of the present invention, areceiver filter is provided that reduces interfering signals from sources external to the wirelesscommunication base station.

In accordance with a preferred embodiment of the present invention, the pluralityof active radiator modules are stacked to form an active antenna having desired gain and beamshape determined by the beam forming network. The modules may be stacked in a vertical array, aplanar array or a circular array, for example.

There is also provided in accordance with a preferred embodiment of the presentinvention, a method for mitigating a fading of signals on a forward link of a CDMA wirelesssystem, the method including splitting a transmission signal to a plurality of transmitter antennas, 4 introducing a delay that is longer than a CDMA chip in a transmit chain of the antennas relative to a first of the antennas, transmitting the signals by all the antennas, receiving the signals withdifferent correlators, and combining the signals, thereby mitigating a fading of the signals.

Preferably each the antenna transmits with approximately equal coverage.

In accordance with a preferred embodiment of the present invention, the step of transmitting comprises transmitting from a plurality of spaced antennas.

Additionally in accordance with a preferred embodiment of the present inventionthe step of transmitting comprises transmitting from a plurality of antennas that transmit atdifferent polarization.

Further in accordance with a preferred embodiment of the present invention thestep of combining comprises combining with natural multipath signals.

There is also provided in accordance with a preferred embodiment of the presentinvention, a modular dual polarized base station antenna system including a plurality of pairs oforthogonal polarization antennas, wherein one of the pairs is polarized at ±45° and another of thepairs is H-V polarized. Preferably a pair of transmit antennas are polarized at ±45°, and a pair ofreceive antennas are H-V polarized. Alternatively all pairs of antennas may be H-V polarized.Preferably each antenna is fed by a separate amplifier.

In accordance with a preferred embodiment of the present invention at least oneisolation structure is provided for increasing isolation between the antenna pairs.

There is also provided in accordance with a preferred embodiment of the presentinvention, a method for modular dual polarized base station transmission and reception, themethod including transmitting with a pair of transmit antennas polarized at ±45°, and receivingwith a pair of receive antennas that are H-V polarized. Alternatively all pairs of antennas may beH-V polarized.

In accordance with a preferred embodiment of the present invention the transmitsignals are split and weights of polarization are applied at a base station. Alternatively, weights ofpolarization are applied by control of amplifier gain. The weights may be applied at RF, IF orbaseband frequencies.

There is also provided in accordance with a preferred embodiment of the presentinvention, a polarization diversity and matching system for cellular radio, including a dualpolarized antenna pair at a base station, each antenna including an appropriate receive channel,and a signal combining and control circuitry that adds polarization diversity to a base station receiver. 5

In accordance with a preferred embodiment of the present invention, the circuitryis characterized by two time constants, wherein a fast circuit adapts to fading signals on areceived reverse link and changes weights of two receive antennas, and a slow circuit followsphysical movements of a mobile station and averages fading of a received signal.

Additionally in accordance with a preferred embodiment of the present invention,the slow circuit is driven by information from received signals and matches transmitted signalpolarization to that of an incoming signal.

Further in accordance with a preferred embodiment of the present invention, asignal from two receiving antennas are weighed by weights controlled by the signal combining andcontrol circuitry, the weights being fed into a transform circuit that transforms the weightsaccording to polarizations of the transmitting antennas and differences in gain.

There is also provided in accordance with a preferred embodiment of the presentinvention, a method for increasing transmission gain to a mobile station of a mobilecommunications system, including substantially simultaneously transmitting from two transmitantennas so as to form a radiation pattern that is characterized by a plurality of radiation lobes,each lobe being characterized by a width inversely proportional to a distance between theantennas, an amplitude of the lobes being bound by the radiation pattern of the antennas,determining a transmission direction to a mobile station, and aiming the pattern so as to produce amaximum in the transmission direction, thereby increasing transmission gain to the mobile station,and reducing scattering into the mobile station from foreign objects.

In accordance with a preferred embodiment of the present invention, the step ofdetermining a transmission direction comprises amplifying and filtering a signal from eachantenna, splitting the signals, and changing a phase of the signals relative to one another so as todetermine the direction. Alternatively, the step of determining a transmission direction comprisesextracting direction information from a receive diversity control for a given antenna channel andcorrecting for a difference in frequency.

There is also provided in accordance with a preferred embodiment of the presentinvention, apparatus for increasing transmission gain to a mobile station of a mobilecommunications system, including two transmit antennas positioned together with a pair ofdiversity receive antennas, wherein the two transmit antennas transmit substantiallysimultaneously so as to form a radiation pattern that is characterized by a plurality of radiationlobes, each lobe being characterized by a width inversely proportional to a distance between thetransmit antennas, an amplitude of the lobes being bound by the radiation pattern of the transmitantennas, and aiming apparatus that aims the pattern so as to produce a maximum in a 6 transmission direction to a mobile station, thereby increasing transmission gain to the mobilestation, and reducing scattering into the mobile station from foreign objects.

In accordance with a preferred embodiment of the present invention, the transmitantennas and the receive antennas comprise a plurality of active radiator module arrays.

Additionally in accordance with a preferred embodiment of the present invention,the transmit antennas are spaced by a distance as required to avoid correlation between fading ofsignals from remote mobile stations within a coverage area.

Further in accordance with a preferred embodiment of the present invention, theapparatus includes a Multiple Carrier Linear Power Amplifier (MCLPA). Preferably the MultipleCarrier Linear Power Amplifier (MCLPA) is located at one of the antennas and connected theretowithout a cable.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will be understood and appreciated from the followingdetailed description, taken in conjunction with the drawings in which:

Fig. 1 is a simplified schematic illustration of a modular base station, constructedand operative in accordance with a preferred embodiment of the present invention;

Fig. 2 is a simplified schematic illustration of a base station with an RF section of asecond generation base transceiver subsystem (BTS);

Fig. 3 is a simplified block diagram illustration of an active radiator module,constructed and operative in accordance with a preferred embodiment of the present invention;

Fig. 4 is a simplified block diagram illustration of an active radiator module forCATV infrastructure based remote microcells, constructed and operative in accordance with apreferred embodiment of the present invention;

Fig. 5 is a simplified block diagram illustration of an active radiator module remotemicrocell via fiber, constructed and operative in accordance with a preferred embodiment of thepresent invention;

Fig. 6 is a simplified block diagram illustration of an active radiator module basedhigh gain antenna array, constructed and operative in accordance with a preferred embodiment ofthe present invention;

Fig. 7 is a table comparing the transmission path power budget of a secondgeneration prior art BTS and that of an active radiator module array, constructed and operative inaccordance with a preferred embodiment of the present invention;

Fig. 8 is a simplified block diagram illustration of a modular design of the active radiator module, constructed and operative in accordance with a preferred embodiment of the present invention;

Fig. 9 is a simplified illustration of mechanical structure of the active radiatormodule, constructed and operative in accordance with a preferred embodiment of the presentinvention;

Fig. 10 is a simplified block diagram illustration of one module of an activeradiator module constructed and operative in accordance with a preferred embodiment of thepresent invention;

Fig. 11 is a simplified block diagram illustration of a stack of modules of activeradiator modules constructed and operative in accordance with a preferred embodiment of thepresent invention;

Figs. 12, 13 and 14 are simplified illustrations of three different arrays of stacks ofmodules, constructed and operative in accordance with three preferred embodiments of thepresent invention;

Fig. 15 is a simplified illustration of a transmission diversity system in a forwardlink of a CDMA base station, constructed and operative in accordance with a preferredembodiment of the present invention;

Fig. 16 is a simplified illustration of a modular dual polarized base station antennasystem, constructed and operative in accordance with a preferred embodiment of the presentinvention;

Fig. 17 is a simplified illustration of a polarization diversity and matching systemfor cellular radio, constructed and operative in accordance with a preferred embodiment of thepresent invention;

Figs. 18, 19 and 20 are simplified illustrations of transmission gain with twoantennas in accordance with a preferred embodiment of the present invention; and

Appendix A is a description of active radiator modules, constructed and operativein accordance with other preferred embodiments of the present invention.

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

Reference is now made to Fig. 1 which illustrates a modular base station,constructed and operative in accordance with a preferred embodiment of the present invention. Acombined signal is transmitted in low power through a cable to a mast, where it redistributes to aplurality of active radiator modules. The number of active radiator modules needed is a function ofboth the total effective radiated power (ERP) and gain required. The receive chain includes an LNA x

in each element, which reduces the noise figure of the system. The same active radiator module can serve in microcells that need small power and low gain antennas.

Reference is now made to Fig. 2 which illustrates a base station with an RF sectionof a second generation base transceiver subsystem (BTS) and an active radiator moduleconstructed and operative in accordance with a preferred embodiment of the present invention.The single channels are combined, after preamplification and channel filtering, and then feed aMulti Carrier Linear Power Amplifier (MCLPA). The combined signal is then band-pass filtered,diplexed and run through a high power, low loss cable, to the antenna array. A disadvantage of prior art second generation BTS is, inter alia, that the complexcomprising the cable and antenna array, serving both transmit and receive signal, is required to beextremely linear and not to generate IMD (Intermod products) higher than about -135 dBc, whichputs a high stress on the antenna and the cable connections. The diplexer and band pass filtersneed about 100 dB of isolation between transmitter and receiver frequencies. The cost of thisarchitecture is a power loss of 3 to 5 dB in the filters, diplexer and cable, that has to becompensated by a high power MCLPA and all its supporting equipment. The cable loss degradesthe noise figure on the receive side.

In the present invention, the MCLPA, high power cable, diplexer and broadbandsuperlinear antennas, and LNA are all replaced by an active radiator module. The active radiatormodule is mounted on the mast and comprises a low power PA, an elemental radiator (dipole or apatch) and a corresponding receive element. The active radiator module performs amplification atlow level and combines the power in the air, uses two narrow band antennas for transmit andreceive, thus reducing the linearization and structural requirements of the antennas, and amplifiesthe received signal at the antenna terminal with no additional loss. The cables connecting theactive radiator module and the BTS are simple and not sensitive to loss, and may be extended asneeded.

Reference is now made to Fig. 3 which is a simplified block diagram illustration ofan active radiator module forming part of the apparatus of Fig. 2. The active radiator moduleincludes two separate transmit and receive antenna elements. This obviates the need for adiplexer, with the associated cost, power loss and occupied volume. Each antenna is preferablynarrow banded, typically covering 12.5 MHz (<2%). A separation of preferably approximately 45MHz provides about 20 dB isolation. Further isolation ( up to 85 dB) is provided by the filters onthe receiver and the transmitter channels. The transmitter amplifier is low power, 2 W or .2 W,depending on the application. A LNA follows the filter on the receiver channel. 9

Reference is now made to Fig. 4 which illustrates an active radiator module forCATV infrastructure based remote microcells, constructed and operative in accordance with apreferred embodiment of the present invention. The basic active radiator module is preferablycombined with a CATV up/down converter module to establish the CATV infrastructure basedremote microcells. This special application active radiator module will make use of the existingCATV network as an RF trunk for remote RF Microcells. Such an existing CATV network is inuse in U.S. markets with a great cost and capacity advantage. A similar product is being offeredby Lucent Technologies.

The CATV up/down converter module input is preferably a coaxial cableconnected to the CATV network and carrying the CATV standard forward link (typically 450-650 MHz) and reverse link (typically 5-52 MHz). A bandwidth of 10 MHz for each forward andreverse links is preferably dedicated to cellular active radiator module use.

The CATV diplexer within the converter separates the transmitter and receiversignals. These are then converted to the appropriate cellular frequencies. Each of these convertersincludes a mixer, phased locked oscillator and a band pass filter to eliminate image and lowfrequencies. The up/down converter module is attached directly to the active radiator module inthis application.

Reference is now made to Fig. 5 which illustrates an active radiator module remotemicrocell via fiber, constructed and operative in accordance with a preferred embodiment of thepresent invention. A fiber/RF converter module is attached to the basic active radiator module forfiber-optics trunking for remote active radiator module microcells. RF trunking via fiber is anefficient method for microcells layout, proposed for both in-buildings and outdoors microcelldistribution. The fiber/RF transducer module preferably includes both transmitter fiber/RFconverter and receiver RF/fiber converter within the same module. The input to this module ispreferably either one fiber carrying both transmitter and receiver signals, split and filtered withinthe module, or two separate fibers, depending on fiber infrastructure. The fiber/RF convertermodule is attached directly to the active radiator module.

Reference is now made to Fig. 6 which illustrates an active radiator module basedhigh gain antenna array. Large cellular cells require both high ERP (Effective Radiated Power)and antenna gain. Arrays composed of active radiator module elements provide both effectively,at lower BTS cost, and higher flexibility and reliability. The ERP generated by a linear arraycomposed of N active modules, each transmitting p Watts, is N2p. There is no additional loss,otherwise included in the link budget due to the BFN (Beam Forming Network), cable anddiplexer. io

Reference is now made to Fig. 7 which is a table comparing the transmission pathpower budget of a second generation prior art BTS and that of an active radiator module array,constructed and operative in accordance with a preferred embodiment of the present invention. Itmay be appreciated that a 10 element array, a 100 W MCLPA, with the associated high powercable and diplexer, may be replaced by 10 active radiator modules, each transmitting 2 W. Asimilar advantage is obtained on the receive path.

Reference is now made to Fig. 8 which illustrates a modular design of the activeradiator module in accordance with a preferred embodiment of the present invention. Eachmodule can be attached to other modules to establish a new product matched to specific customerrequirements.

The active radiator module preferably comprises five basic building blocks and theintegrating enclosure. The basic building blocks are: 1. transmitter amplifier 2. receiver amplifier 3. transmitter/receiver band pass filters 4. transmitter/receiver antenna element 5. power supply

Before describing the amplifier of the active radiator module of the presentinvention, a brief discussion of the prior art will now be presented. Generally in the prior art,stringent intermod products (IMP) specifications are imposed on the BTS transmission, in orderto avoid interference to its own receivers and to adjacent cells and systems. These impose linearityrequirements on the transmission chain beyond the channel filters. The MCLPA for a large cell isthus specified not to exceed -70 dBc IMP. These constraints do not apply for a single channelamplifier, and are relaxed for microcells, where the dynamic range of the cell is reduced by over30 dB. The IMP requirements for CDMA systems are less stringent. A class A amplifier with proper backoff (3 to 5 dB) may serve the requirements forCDMA microcells and for other systems’ low capacity microceils and cells (“minicells”). Higherlinearity requires linearization techniques. Pre-distortion results in 7 to 10 dB higher 3rd ICP (thirdorder intercept point) and enables 10 to 20 dB lower IMP. A cost factor of 10 is considered todaypractical compared to the class A amplifier, owing to the hybrid design of the latter, as comparedto discrete components and manual tuning of the pre-distorted amplifier. The cost may be reducedby resorting to a similar technology, justified for large quantities. Further linearization requiresfeed-forward techniques as used in the high power MCLPA. It is expected that these expensivetechniques will not be needed in any of the active radiator module applications. 11

The amplifier of the active radiator module of the present invention will now be described. A class A amplifier, operating at 3 to 5 dB backoff, may be used for the first generation of active radiator module. Such an amplifier offers low cost and a short development time. The amplifier accommodates the CDMA microcells and minicells and low capacity microcells and minicells for other systems, which constitute a major portion of the market.

The transmitter amplifier preferably comprises two stages. The first stage ispreferably a monolithic silicon class A gain stage. The second stage is preferably a hybridpackaged power amplifier. The amplifier with all of its matching and biasing networks arepreferably assembled using SMT technology on a RF printed board within the transmitteramplifier enclosure.

Typical transmitter amplifier specifications are presented here for purposes ofdescription of best mode, but the present invention is not limited to these values.

Frequency band.....................870-895 MHz (AMPS, TDMA, CDMA) 935-960 MHz (GSM)

Output power (average)...........2 w, saturated for AMPS

1 w, for TDMA GSM0,6 w, for CDMA 1 dB compression........................32-33 dBm 3rd order intercept point.............+43 dBm min.

Two tones IMP...........................-40 dBc for 0.2 w per tone -46 dBc for 0.1 w per tone

Gain..........................................30 dB, @ small signal

Gain flatness.............................+/- 0.5 dB over any 1.25 MHz

Absolute phase variation............+/- 5° over any 1.25 MHz AM/PM conversion...................025°/dB Max up to 3 dB below 1 dB cp

Noise figure..............................7.5 dB Max

Input VSWR.............................1.5 : 1 @ 50 ohm system output VSWR............................1.3 : 1 @ 50 ohm system

DC supply voltage......................8 to 18 volts DC DC supply current.....................4.3 to 1.9 Amp

Operating temperature...............-20° to +60 °C

The receiver amplifier within the active radiator module is intended to ensure thatthe base station sensitivity will not be degraded because of long coaxial cables losses or othermedia losses and noise add in between the antenna element and base station front end. 12

The receiver amplifier preferably has enough gain, low enough noise figure, highenough compression and intercept points to eliminate sensitivity, inter-channel interference andnon-linear multi-channel distortion degradation.

Typical receiver amplifier specifications are presented here for purposes ofdescription of best mode, but the present invention is not limited to these values.

Frequency range................................................800-950 MHz

Noise figure......................................................3.5 dB

Gain...................................................................30 dB

Input 1 dB compression point............................-10 dBm

Input 3 rd order Intercept Point............................0 dBm

Class of operation..............................................A

Voltage supply..................................................+8v regulated

Current requirement..........................................150 mA

Operating Temperature......................................-20°C to +60°C

Technology.......................................................SMT of MMIC

Transmitter and receiver filters of the active radiator module establish, togetherwith transmitter/receiver antennas separation, a diplexer which isolates transmitter and receiversignals from each other. Specifications for transmitter and receiver filters are directly driven fromperformance requirements of active radiator module per system and application.

Typical specifications are presented here for purposes of description of best mode,but the present invention is not limited to these values.

Noise power input to ARM @ receiver band................................-155 dBm/Hz

Transmitter channel noise figure............................................................10 dB

Transmitter channel gain........................................................................35 dB

Channel Band Width (for AMPS, TDMA)...............................30 KHz

Transmitter output noise @ transmitter antenna over channel B.W...................-74 dBm (-155 +10+35+35=-74)

For the receive channel: MDS( Minimum Detectable Signal) for AMPS, TDMA)..........-125 dBm

Input 3rd Intercept Point..........................................................0 dBm

Channel B.W (for AMPS, TDMA)...........................................30 KHz

Receiver Noise Figure......................................................................4 dB

Receiver noise power at ARM input................................................-135 dBm (KTBF=-174+35+4=-135) 13

The transmitter filter has two roles within the active radiator module. The first is toreduce transmitter wide band noise in a receiver band. The second is to reduce spurious signalswhich might interfere with a receiver channel of the same cell or other cells or other systems. Themore demanding requirement is the first one and it dictates the transmitter filter performance andthus transmitter filter structure.

In order for the transmitter noise and leakage into the receiver channel input to belower than the receiver noise floor, an isolation of 60 dB (-74+135) is required. 20 dB of therequired isolation is attributed by transmitter/receiver antenna isolation and the other 40 dB plus10 dB of safety margin , are given by receiver band rejection of the transmitter filter. The samereasoning holds for CDMA systems where the values differ but the ultimate results hold.

The receiver filter has two roles within the active radiator module. The first is toreduce the transmitter signal to a level where interfering intermod products are not generated inthe receive chain, and the receiver amplifier is not desensitized by saturation. The other purposeof the receiver filter is to reduce interfering signals from other base stations and other systems.The more demanding requirement is the first one and it will dictate the receiver filter performanceand thus the filter’s structure.

In order for the transmitter leakage not to interfere with received signal, it shouldbe kept at a much lower level than receiver channel compression for systems with no AGC orwhen AGC is at minimum. For example, if the 1 dB compression point at the receiver antennaterminal is -60 dBm, the transmitter leakage is preferably below -70 dBm. For transmitter averageoutput power of +33 dBm and transmitter/receiver antenna isolation of 20 dB, the receiver filterrejection of transmitter band is preferably 85 dB. The same reasoning holds for CDMA systemswhere the values differ but the ultimate results hold.

Typical specifications for the active radiator module filters are presented here forpurposes of description of best mode, but the present invention is not limited to these values.

Transmitter Filter:

Pass band ...................................10 MHz over 870-895 MHz (AMPS) or 935-960 MHz (GSM)

Rejection ..................................-50 dB @35 MHz below pass band -40 dB @ 35 MHz above pass band

Insertion loss................................-1 dB max. @ pass band

Ripple within band........................0.4 dB max. over any 1.25 MHz band

Group delay variation 0.5 nsec max. over any 1.25 MHz

Receiver Filter 14

Pass band ...................................10 MHz over 825-850 MHz (AMPS) or 890-915 MHz (GSM)

Rejection ..................................-85 dB @35 MHz above pass band % -40 dB @ 35 MHz below pass band

Insertion loss................................-1 dB max. @ pass band

Ripple within band........................0.4 dB max. over any 1.25 MHz band

Group delay variation 0.5 nsec max. over any 1.25 MHz

Reference is now made to Fig. 9 which illustrates mechanical structure of an activeradiator module 100 in accordance with a preferred embodiment of the present invention. Activeradiator module 100 preferably includes a housing 102 typically constructed of aluminum. Aplurality of tuning elements 104 and I/O connectors 106 are preferably mounted on outsidesurfaces of housing 102. Input from transmitter and receiver antenna elements 108 and 110,respectively, is fed to the I/O connectors 106 and output connections are preferably directly madewith the amplifier’s circuit boards (not shown). Disposed inside housing 102 are transmitter andreceiver filters 112 and 114, respectively. Each of the transmitter and receiver filters 112 and 114preferably includes a 6-coaxial-resonators elliptic filter in combine structure.

Both transmitter and receiver antenna elements 108 and 110 are preferably printedpatch elements. Transmitter and receiver elements 108 and 110 are preferably printed on the samebase material (typically polyurethane material) and covered by a sheet of epoxy-fiberglass or otherprotective cover that withstands the environment, including UV radiation. Both elements 108 and110 are preferably designed to ensure the required isolation between the elements. The design ispreferably compatible with array-stacking of elements for a high-gain antenna, as will be describedfurther below.

Typical specifications are presented here for the active radiator module antenna forpurposes of description of best mode, but the present invention is not limited to these values.Frequency band..................Transmitter: 870-895 MHz for AMPS, TDMA, CDMA

935-960 MHz for GSM

.................Receiver: 825-850 MHz for AMPS, TDMA, CDMA

890-915 MHz for GSM

Transmitter/receiver elements isolation.............20 dB min. VSWR........................................1.4 : 1 Max

Size............................................150x150 mm

An active radiator module power supply preferably supplies all DC powerrequirements of the transmitter and receiver amplifiers and includes all protection means needed 15

for a tower top mounted device. Since the active radiator module power supply is preferablymounted on top of the antenna tower and cable connecting the base station and the active radiatormodule does not have a prefixed length, a DC-DC converter is needed within the power supply.DC supply is preferably through transmitter or receiver coaxial cables, which means a BIAS-Tshould be implemented within the active radiator module power supply. The DC supply source ispreferably within the base station. This way of DC supply is convenient for the modular approachwhere each module (CATV converter or Fiber/RF converter) has an independent power supply,all consuming DC power from the same source through connecting coaxial cables.

Typical specifications are presented here for the power supply for purposes ofdescription of best mode, but the present invention is not limited to these values.

Input voltage..............................18-32 V DC

Output voltages............................15V DC @ 2 Amp (or other voltage with same power) .............................8 V DC @ 0.2 Amp

Operating temperature..................-20°C to +60°C

Input connection...........................Through an internal BIAS-T

The integrating enclosure of active radiator module is based on atransmitter/receiver filters block. This block occupies the main volume of the active radiatormodule. The mechanical structure is preferably divided into two main mechanical parts: the mainblock of the filters with printed antenna and radome mount as one part and filters cover withtransmitter amplifier, receiver amplifier and power supply compartments as second part. Bothmechanical parts are preferably made of die cast aluminum and screwed to one another withsealing conductive O-ring in between the parts. Overall size of active radiator module structure ispreferably around 150 x 150 x 150 mm.

The active radiator module aluminum structure is preferably designed to dissipateheat from the transmitter and receiver amplifiers and power supply. Overall heat dissipated withinthe active radiator module is about 30 W and the temperature rise above ambient temperature isapproximately 10 °C or less.

The antenna radome is preferably at the front of the active radiator module, whilethe transmitter and receiver connectors are preferably on the rear side. Several active radiatormodule units may be interconnected to form an array for the higher gain and higher powerantennas.

As mentioned above, the active radiator modules are compatible with array-stacking of elements for a high-gain antenna. Reference is now made to Fig. 10 which illustrates a 16 block diagram of one module of an active radiator module constructed and operative in accordance with a preferred embodiment of the present invention. The active radiator module preferably includes a power amplifier, band pass filter, diplexer and LNA in conjunction with a radiating element, such as an antenna.

Reference is now made to Fig. 11 which illustrates a block diagram of a stack ofmodules of active radiator modules constructed and operative in accordance with a preferredembodiment of the present invention. A low power cable is used with the transmit beam formingnetwork.

Reference is now made to Figs. 12, 13 and 14 which illustrate three differentpossible arrays of stacks of modules, constructed and operative in accordance with a preferredembodiment of the present invention. Fig. 12 illustrates a vertical array, Fig. 13 illustrates a planararray and Fig. 14 illustrates a circular array. It is appreciated that other configurations are possiblein the scope of the invention. Alternatively, in the embodiments of Figs. 10, 11, 12, 13 and 14,instead of employing a diplexer, two separate antennas for transmit and receive may be used.

The following list summarizes some of the advantages of the active radiatormodule of the present invention: • A common module for many cellular applications • A major cost saver for large cells (reduces the power requirement by 3 to 6 dB as compared toMCLPA at the base station (BS) compartment, reduces the need for high power low loss cable,and enables longer distance between the antenna and the BS). • Takes a substantial “bite” out of the BS market, while avoiding the need for expensivetechnology licenses associated with the BS production. • Reduces the BS noise figure by eliminating the cable losses. • Reduces antenna Intermod Products (IMP) by separating the transmitter and receiver antennasand by eliminating cable connections. • A failure of a module in an array does not cause a catastrophic damage to the BS, but only agraceful degradation. • A building block for multiple-beam and smart antennas.

In accordance with another embodiment of the present invention, a method andapparatus are provided whereby a fading of a mobile channel can be mitigated on a forward linkof a CDMA wireless system, by transmission from two or more spaced antennas, or from twoantennas that transmit at different polarization. Part of this concept has been applied to distributedantennas, as described in United States Patent 5,280,472, issued Jan 18, 1994, the disclosure ofwhich is incorporated herein bv reference. The present invention is different, however, in that it 17

applies to the transmit antennas only, particularly to large base stations, and for cellular service, while the referred patent applies to PBX indoors, and a serial chain of elements, each serving both for transmit and receive.

By way of background, it is noted that the forward and the reverse links in mostcellular systems are transmitted at different frequencies (what is called FDD - frequency divisionduplex). The typical spacing for the 800 or 900 MHz systems is 45 MHz, which is larger than thecoherence bandwidth of the typical terrestrial cellular communications channel. There is notenough correlation between the fading in both channels, and the fading information from thereverse link cannot be used to control a diversity transmission in the forward link. A unique feature of the CDMA system is the “rake receiver”, which consists of anumber of digital correlators that enable the reception of different transmissions, and in particular,the same transmission with a different delay. The receivers in the IS 95 system, for example, areequipped with 4 such correlators (called “fingers”) that, in the intended mode of operation,search for delayed multipath transmissions. This feature can be used for intentionally delayedtransmission from two or more antennas whose fadings are not correlated, e.g., antennas that aresufficiently spaced apart, or that transmit with different polarization.

Reference is now made to Fig. 15 which illustrates a transmission diversity systemin a forward link of a CDMA base station, constructed and operative in accordance with apreferred embodiment of the present invention. The embodiment is for a base station builtaccording to the Interim Standard IS 95, and related standards, but is not limited to thesestandards only. The transmission signal is preferably split to the number of antennas to be used fortransmission, typically two antennas 131 and 133. A delay that is longer than the CDMA chip (aterm known to those familiar with CDMA technology) is inserted in the transmit chain of each ofthe other antennas relative to the first antenna. This delay may be inserted in the RF chain, by adelay line 301 in Fig. 15, (a term known to those familiar with FRF technology) or in the digitalbase band. With the IS 95, the required delay is higher than 1.5 microseconds. The signal is thentransmitted by all antennas, each having about the same coverage. These signals are then receivedby the mobile station with different “fingers” and optimally combined, as with natural multipathsignals (a process known to those familiar with CDMA technology). The diversity gain thusobtained may be significant.

The typical delay profile depends on the environment. In urban and suburbanenvironments, the first delay cluster may extend to about 2 microseconds. An optimal delay ispreferably selected for the transmission in order to minimize interference from natural multipathsarriving at the same time. is

In accordance with another embodiment of the present invention, dual polarized active antennas are provided for BS of mobile communication systems. This dual polarized pair for transmit, and one for receive, allow the BS to use polarization diversity on receive, and polarization matching on transmit.

Diversity is a powerful way to mitigate the multipath in mobile communicationssystems. Most of the cellular systems in use today employ antenna diversity in the base station forreceiving the reverse link - transmission from the mobile station (MS). However, the mobilestations do not include multiple antennas and diversity receivers for reasons of cost andcomplexity. Diversity in use by most is space diversity, whereby the two (or more ) receiveantennas are spaced apart in order to reduce the correlation between their fading.

The correlation between orthogonally polarized signals is also low, particularly forurban and indoors propagation, and polarization diversity is also an efficient mitigation meansagainst channel fading. One very important advantage of polarization diversity is the compactnessof the antenna arrangement involved.

Dual polarized antennas may also be used at the BS for polarization matching ofthe forward link (transmission from the BS to the MS), as described further hereinbelow. The useof modular active antennas in the BS is also described further hereinbelow.

Reference is now made to Fig. 16 which illustrates a modular dual polarized basestation antenna system, constructed and operative in accordance with a preferred embodiment ofthe present invention. The system preferably includes two pairs of orthogonal polarizationantennas. The embodiment of Fig. 16 exhibits a compact arrangement, whereby one pair 202 ispolarized at ±45° while another pair 204 is H-V (horizontal-vertical) polarized. This particulararrangement has a relatively high isolation between the transmit and receive pairs, and each pairmay be tuned to its desired frequency range. Alternatively all pairs of antennas may be H-Vpolarized. Other arrangements are possible, e g. patches.

The transmit antennas are typically polarized at ±45°. Alternatively they may be H-V polarized. In one embodiment, each antenna is fed by a separate amplifier. The transmit signalsplitting and weights of each polarization are performed at the base station at a low RF level, orIF or at the baseband, thus avoiding RF losses. One other alternative is to apply the weights bycontrol of the amplifiers gain. The use of two amplifiers allows for the summation of theirtransmit power in the air, for any polarization determined by the polarization matching circuit.

The receive antenna is typically polarized at H-V. In one embodiment, eachantenna preferably feeds an LNA (Low Noise Amplifier). The weights to each signal are appliedafter amplification, at the BS. in RF, IF or baseband frequencies. 19

Each active radiator module unit consists of two dual polarization antenna pairs, as exemplified in Fig. 16. Isolation structures, as shown by lines 206, are used in one embodiment for increasing the isolation between the antenna pairs.

In accordance with another embodiment of the present invention, a method andapparatus are provided whereby a mobile cellular communications system can mitigate the fadingenvironment and compensate for the loss due to mismatch of the polarization of the BS and themobile terminals.

The electromagnetic radiation is polarized, and allows for two orthogonalpolarization states. Any antenna cannot be matched simultaneously to both polarizations. Thepropagation of the signals through a nonhomogeneous medium may transfer part of the signals toan orthogonal polarization. This is the case for terrestrial communications, for example, and inparticular in urban areas, where the signals encounter multipaths from objects on the way. Thetransfer of polarization has been found to be typically -6 to -10 dB in rural areas, and much more(-3 dB or higher) in urban areas and indoors. The orthogonal polarization components have beenfound to have an independent fading pattern, with correlation of less than 0.6.

Cellular systems operate traditionally in vertical polarization. Most of the mobileterminals, mounted on vehicles, have vertically polarized antennas. The polarization of hand-heldterminals is variable, however, and depends on the particular terminal, its orientation relative to ahead and the base station. There is an a priori mismatch of the polarization between the hand-heldterminal and the BS antenna.

The forward (from the BS) and the reverse (from the MS) transmissions do nothave correlated fading, because of the frequency difference between them. Diversity schemes,used for mitigating the fading, are therefore applicable on receive only, due to lack of informationto control transmission so as to reach the other receiver without fade. In most of the present daysystems, only the base stations are equipped with more than one antenna for receiving the signalfrom each mobile station. Mobile stations that are equipped with only one antenna are unable touse antenna diversity for reception.

Reference is now made to Fig. 17 which illustrates a polarization diversity andmatching system for cellular radio, constructed and operative in accordance with a preferredembodiment of the present invention. A dual polarized antenna pair at a base station, with an appropriate receive channelfor each, and a signal combining and control circuitry, adds polarization diversity to a base stationreceiver, by itself or in addition to other diversities. In the present invention the adaptivecombining control circuit has two time constants. First, a fast circuit adapts to fading signals on a 20

received reverse link and changes weights of two receive antennas. Second, a slow circuit followsonly the physical movements of the MS, but averages fading of the received signal. This control isapplied to a dual polarization transmission antenna pair, with their appropriate transmissionchannels. This slow control, driven by the information from the received signals, matches thetransmitted signal polarization to that of the incoming signal. This enhances the forward linksignificantly in situations of large polarization mismatch, that may be frequent in normaloperation.

The signal from the two receiving antennas are weighed by weights 150, controlledby the measurement and control circuit. This can be applied in RF, IF or at the baseband. Theseweights are also fed into a transform circuit, that transforms the weights according to the differentpolarization of the transmitting antennas (e.g. tilted linear, or circular) and differences in gain.The low-pass filter averages the fast control variations and responds only to the slow variationsresulting from the physical attitude changes of the MS. Weights 152 may be applied to thetransmitted signal at the base band, at the IF, prior to power amplification, as shown, or afteramplification, or applied to the amplifiers gain control. The same antennas may also be used forboth reception and transmission. In such a case the signals are separated by a diplexer.

In accordance with another embodiment of the present invention, a method andapparatus are provided whereby a mobile communications system can enhance the transmittedsignal to each mobile station(MS) while reducing the multipath fading, and the interference toother MS.

Transmission in a terrestrial environment encounters multipaths due to scatteringfrom various objects. The multipath components, arriving at the receiver, interfere with each otherto form typical fading of the signal, when either transmitter, receiver or scattering obstacles are inmotion. This is one of the most detrimental effects on the communications in this service.Diversities are employed to mitigate the fading, whereby propagation routes which have anindependent fading (non-correlated fading) are received separately and then combined in a waythat reduces the variations in the signal. Base stations in the cellular service typically employ tworeceive antennas, spaced apart so as to have no correlated reception from the mobile stations intheir coverage area. Diversity on the forward link (transmission from the BS) is not implementedin most systems because of the excess complexity needed in the MS by introducing two antennasand the accompanying circuitry. In transmission diversity from the BS, two or more transmitantennas are positioned far apart to avoid correlation, and their transmitted signals are controlledin amplitude and phase in correspondence to the output of the adaptive control circuitry of thereceiving antennas. Such an arrangement does not apply to the cellular systems where the forward 21

and the reverse links transmission have different frequencies, far enough to decorrelate the fading between the two links.

Narrow beam transmission from the BS has been proposed, in order to increasethe signal strength received by the MS, and reduce the interference to other MS. Narrow beamtransmission also reduces the multipath interference to the BS by limiting the illuminated areaaround the BS. Such a narrow beam has to track the desired MS direction. Algorithms proposedfor this task make use of the direction information obtained from the diversity antennas. A BS thatcommunicates with many mobile stations needs to form simultaneously many narrow beams so asto optimize the transmission to each station. Such an arrangement requires a multibeam antennaarray which is a relatively complex system.

Reference is now made to Figs. 18, 19 and 20 which illustrate transmission gainwith two antennas in accordance with a preferred embodiment of the present invention.

Simultaneous transmission from two antennas forms a radiation pattern that ischaracterized by many radiation lobes, the width of each is inversely proportional to the distance(in wavelengths) between the antennas. The amplitude of these lobes is bound by the radiationpattern of the individual antennas. Such a pattern, when aimed in a way to produce a maximum inthe direction of a mobile station, has a gain of 2 (3 dB) in that direction. The narrow lobe aroundthis maximum also reduces the scattering into the mobile station from the objects in the area thatis now illuminated less. The typical signals arriving at the mobile station from the scatteringobjects tend to diminish roughly in proportion to the second power of the distance of thesescatterers from the mobile stations. Scatterers that are illuminated by the adjacent lobes contributemuch less to the fading at the mobile stations. The contribution of the multilobe antenna to themitigation of fading, when its radiation is properly aimed, is significant and comes close to that ofa multibeam array whose size matches the distance between the antennas. This is illustrated in Fig. 18.

The interference to other mobile stations is also reduced. Though some mobilestations, positioned at the peaks of other radiation lobes, receive the same signal strength as thedesired mobile stations, other mobile stations within over 50% of the coverage area receivesignals that are at least 3 dB lower, and over 10 dB lower within 20% of the area.

The application of such an arrangement is especially advantageous when thetransmit antenna pair is positioned together with the diversity receive pair, on the sameinstallation. A preferred embodiment is a pair of active radiator module arrays, each provided withreceive and transmit antennas and amplifiers.

Reference is now particularly made to Fig. 19. The signal from each channel isamplified, filtered and then split to two active radiator modules. A phase control changes therelative phase of one active radiator module in respect to the other, to form the desired lobedirection. The information for directing the lobe may be achieved in different way. One preferredembodiment is by extracting direction information from the receive diversity control for thatchannel and correcting for the difference in frequency. Phase ambiguity is not important in thiscase, as it is with true direction finding, because the purpose is to point one of the lobes’ peakstoward the MS and it is not important to determine which lobe. There are many estimationalgorithms that can apply.

Referring to Figs. 19 and 20, a pair of transmitting antennas 401 and 408 arepreferably co-located with base station receiving antennas. Antennas 401 and 408 are preferablyspaced by a distance as required to avoid correlation between fading of signals from remotemobile stations within the coverage area (411 in Fig. 20), typically 10 wavelengths. A cable 402 ispreferably provided to carry high power transmit signals (RF). A Multiple Carrier Linear PowerAmplifier (MCLPA) 403 is preferably provided, as well as a cable 404 that carries low powertransmit signals. In one preferred embodiment, MCLPA 403 is located at one of the antennas andconnected thereto without a cable 402, thus avoiding additional losses. A passive combiner 405preferably combines signals from the individual carriers, as is known in the art. A channel filter406 and a single channel low power amplifier 407 (pre-amplifier), both well known in the art, arepreferably provided for the transmitted signal. A signal control element 409 may be implementedin the RF, as shown in the figure, by a phase shifter, as is known in the art. In an alternativeembodiment, signal control element 409 may be implemented in the IF or at the base band. Thelobe of the coverage is referenced by numeral 410, and is created by the interference between thetransmission signals of antenna 401 and 408.

Appendix A further discloses active radiator modules, constructed and operative inaccordance with other preferred embodiments of the present invention.

It is appreciated that various features of the invention which are, for clarity,described in the contexts of separate embodiments may also be provided in combination in asingle embodiment. Conversely, various features of the invention which are, for brevity, describedin the context of a single embodiment may also be provided separately or in any suitablesubcombination.

It will be appreciated by persons skilled in the art that the present invention is notlimited by what has been particularly shown and described hereinabove. Rather the scope of thepresent invention is defined only by the claims which follow: 23

APPENDIX A

ARM ACTIVE RADIATOR MODULEProducts and Applications

June 22 1997

SUMMARY ARM - Active Radiator Module, is a novel approach to cellular base stationsarchitecture, that offers better performance, additional flexibility and reliability, andlower cost. This transmit-receive module is compatible with a variety of cellular basestation configurations, while following the trend of innovation of cellular architecture.Installation of the ARM in existing base stations requires minimum modifications, whileadding to the performance and reliability, and providing the basic structure for a host ofnew features. New base stations built with the ARM benefit the most. ARM is a transmit- receive module that incorporates the transmit chain - including thepower amplifier, filter and elemental antenna, and the receiving chain that includes theelemental antenna, filter and Low Noise Amplifier (LNA) - in a single compact module.The module is self contained, with its power supply and control/ monitoring unit, and isbuilt as a stand-alone unit or for easy stacking in columns, rows or planar arrays withany desirable gain and polarization, and any desirable beam control. This design avoidsthe expensive loss of power between the base station and the antenna - which amountsto up to 10 dB, and increases the sensitivity of the receiver by 3 to 5 dB, thus increasingthe coverage by 40 to 70% and reducing the requirements on power and accessories inthe base station. The redundancy in power amplifiers in the antenna array, eachtransmitting low power, provides an unmatched reliability and graceful degradation. Themodular design allows for site matching and optimization, and for fast introduction ofnew exciting application, from dual polarization to transmit diversity and smart antennas.The dimensions of the ARM matches those ofthe antennas that it replaces.

Summary - The rationale for ARM • Common module to all BS antennas • Flexibility in cell design • Fast introduction of upgrade and of new features 1

• Reduction of Power Amplifier Requirements by 3 to 10 dB • Reduction of BS compartment size • Eliminates the MCLP A Rack • Reduces the Heat Exchange requirements • Reduces power and emergency power needs • No loss limitation on cable • Can remote the BS, ease zoning and reduce rent • Coverage increase by 50 - 70%

• Receiver sensitivity increased by 2 to 5 dB • Reduces non-linearities and intermod products • Eliminates high power connectors • Antennas and amplifiers are an integral unit • Reduces cost • $40,000 to $60,000 saving per cell • Additional savings for the compartment size and location • Increases reliability • No single failure is catastrophic • Unmatched reliability to arrays • Microcells

Modular Microcells with RF, fiber or CATV trunkingCDMA delay distributed antennas • Suits enhancement trends • Multiple beam antenna arrays and smart antennas • Can provide CDMA Transmission diversity • Multiplexed trunking reduces the number of cables 2

PART ONE - THE ARM 1. ARM description 2. ARM configurations 2.3. High gain antenna array 2.4. Dual polarized antenna 2.5. Transmission diversity delay module for CDMA IS95 2.6. Multibeam/ intelligent antenna arrays 2.7. Multiplex trunking for multibeam/ intelligent antenna arrays 2.8. CATV trunking module 2.9. Fiber optics trunking module 10. ARM high-level design and specifications 11. ARM array interface with the base station. 12. ARM-based base station cost comparison 13. ARM array reliability analysis PART TWO ARM system applications 7.1. Polarization diversity and matching with ARM arrays. 7.2. Transmission tracking with ARM arrays. 7.3. Transmission diversity in CDMA with ARM arrays. 7.4. Multibeam and tracking antenna with ARM arrays.

7.5. ARMCell - design guidelines for cell design with ARM 3

1. DESCRIPTION OF THE ARM MODULE 1.1. The Base Transceiver Subsystem (BTS)

The RF section of a second generation of Base Transceiver Subsystem (BTS) is depictedin Fig. 1.1. The single channels are combined, after preampliftcation and channel filtering,and then feed a Multi Carrier Linear Power Amplifier (MCLPA). The combined signal isthen band-pass filtered, diplexed and ran through a high power, low loss cable, to theantenna array. The complex, consisting of the cable and antenna array, serving bothtransmit and receive signals, is required to be extremely linear and not generate IMD(Intermod products) higher then about -135 dBc - which stresses the antenna and thecable connections specifications. The diplexer and Band Pass Filters need about lOOdb ofisolation between Tx and Rx frequencies.

Fig. 1.1: RF section of one sector of Base Transceiver Subsystem (2nd generation)

The cost of this architecture is a power loss of 3 to 10 dB in the filters, diplexer .cableand antenna beam forming network, that has to be compensated by a high power MCLPAand all its supporting equipment. The antenna beam forming network and the cable lossdegrades the noise figure on the receive side by 3 to 5 dB.

The ARM replaces the MCLPA, high power cable, diplexer and broadband superlinearantennas, and LNA, by a module on the mast that amplifies the low level transmissionsignal to a moderate level that is radiated by the elemental antenna that is integrated in themodule. The radiation of all modules in a column array is combined in the air to producethe required ERP. Two narrow band antennas for transmit and receive are integrated ineach, thus reducing the linearization and structural requirements of the antennas andalleviating the need for a diplexer. The received signal is amplified at the antenna terminalwith no additional loss. Since BTS with ARM system performance is not sensitive tocables losses, the cables connecting the ARM and the BTS are inexpensive and may beextended as needed. 4

1.2. Description of the basic ARM

The Active Radiator Module (ARM) block diagram is shown in Fig. 1.2. The ARMincorporates two separate transmit and receive antenna elements. Each antenna isnarrow banded, (<3%). The separation between the Tx and Rx bands provides about20 dB isolation, and obviates the need for a diplexer, with the associated cost, powerloss and complexity. The band pass filters on the Rx and the Tx channels provide theadditional 85 dB isolation required between the links. The Tx amplifier is low power,providing up to 2 Watts of output power. A LNA follows the filter on the Rx channel.Each module has its own power supply and monitoring-and-control function.

Fig 1.2: ARM-Active Radiator Module Block diagram

Figl.3 : Arm unit general view (PCS version) 5

The reliability of a single ARM module is 200,000 hours MTBF. When stacked into anarray, as in most cells, the redundancy of the ARM s in the array provides anunmatched reliability of the order of 109 hours MTBF.

The ARM is provided with either a vertical or a horizontal polarization on eachtransmit and receive antennas, and polarization can be changed by replacing theantenna form-fit module at the ARM front face. This is due to the unique design of theunit.

The modular, self contained, structure and functioning of ARM, its dimensions - thatsuit antenna arraying both in vertical columns and horizontal multibeam arrays, and therobustness of its tuning parameters, offer a unique flexibility for arraying andprocessing the base station radiation elements - by simple engineering and withoutresorting to further development for each new task. This feature is the key foroptimizing the configuration of each cell, to many enhancement features and toemerging “smart antennas”.

2. ARM CONFIGURATIONS 2.1. High gain antenna array 2.2. Dual polarized antenna 2.3. Transmission diversity delay module for CDMA IS95 2.4. Multibeam/ intelligent antenna arrays 2.5. Multiplex trunking for multibeam/ intelligent antenna arrays 2.6. CATV trunking module 2.7. Fiber optics trunking module 2.1. ARM Based High Gain Antenna Array

Large cellular cells require both high ERP (Effective Radiated Power) and antennagain. Arrays composed of ARM elements provide both effectively, at lower BTS cost,and higher flexibility and reliability. The ERP generated by a linear array composed ofN active modules, each transmitting p Watts, relates to N2p (the gain of a linear arrayrelates to the number of elements N. This is multiplied by the number of amplifiers N).There is no additional loss, otherwise incorporated in the link budget due to the BFN(Beam Forming Network), cable and diplexer. Table 2.1 exemplifies the comparison. A100 Watt MCLPA (Multi Carrier Linear Power Amplifier), feeding a 10 elementantenna array through the associated high power cable and diplexer, is compared to anarray of 10 ARMs, each transmitting 2 W.

Table 2.1: Transmission path power budget

Item PA(dBw) Filter Diplexer Cable +BFN Array gain ERP for N elements ERPfor 10elements 2nd gen. BTS 20 -.5 -2 -5 10 LogN 12.5+10 LogN 22.5 dBw ARM array 3+ 10 LogN -.5 0 0 10 LogN 2.5+20 LogN 22.5 dBw 6

The reliability of the transmitter is improved considerably at the same time by theredundancy of the amplifiers. The same is true for the receive chain, where the noisefigure is improved by 2 to 5 dB, which is translated to additional coverage of up to70%.

The ARM is configured to allow stacking in linear and in two dimensional arrays. Twocorporate feeds, for Tx and for Rx, provide the respective beam forming. These maybe custom designed for specific tilts or beam shaping. The array block diagram isshown in Fig, 2.1., and the modular configuration of the ARM stacks is shown inFig.2.2.

Txcoax-

CORPORATEFEEDDIVIDER

A. R. M

TO MODIFIEDBASE STATION

ARM

SECTORIAL

ANTENNA

ARRAY

Rxcoax -

Base Station RF Interface

CORPORATE

FEED

COMBINER I______

FU

A. R. M

________________I

Fig. 2.1.: ARM High Gain Antenna Array Block Diagram

Fig. 2.2.: ARM physical configuration a. Single module b. Active array antenna 7 /~~7. ~7

/._/ / / a

Fig. 2.3 : ARM arrays a - A 4 element columnb - An 8 element columnc - A 8 x 4 planar / multibeam arrayd - A circular array e - An horizontal array backed by a corner reflector ARM column Beam Forming Network

Fig 2.4.: ARM Array corporate feed 2.2. Dual polarized ARM antenna

This dual polarized pair for transmit, and one for receive, allow the BS to usepolarization diversity on receive, and polarization matching on transmit. Theapplications are described in part 2. The ARM can incorporate vertical or horizontalpolarization on both the Tx and the Rx antennas. This is due to the unique design ofthe antennas and their feeds. This flexibility offers a variety of implementations. Fig.2.5a describes the configuration of polarization diversity on Rx, while the Tx isvertically polarized and power-enhanced. A configuration for polarization diversity onRx and on Tx, or polarization matching on Tx, is described in Fig. 2.5b. 8

2.3, Delay Diversity Module for CDMA IS 95

The Delay diversity Module enables the Tx CDMA delay diversity option whererequired . The modules main blocks are the SAW delay line and the amplifier whichcompensates for the delay line insertion loss. The dual bias-T by-passes this module forDC current and M&amp;C signals wile biasing the internal power supply off the main Txcoaxial cable supply. The Delay diversity module is attached at the Tx Beam-FormingNetworks input or at an individual ARM Tx input where applicable. Fig. 2.6. showsmodules block diagram.

An implementation of this module to a CDMA Delay Diversity Distributed Antenna isshown in Fig. 2.7.

Fig.2.6. : Tx Delay Diversity Module Block Diagram 9

Fig.2.7.: CDMA Delay Diversity Distributed Antenna 2.4. ARM for multibeam/adaptive arrays

The capacity, gain and performance of the cellular systems are expected to improve bythe use of “Intelligent antennas” (or “smart antennas”). These are arrays of antennas,as shown in fig.2.4c, d and e, that can form multiple narrow beams, or shape acompound beam, that is adaptive to the teletraffic activity. The intelligence for creatingthe beams is derived from the signals received by the BTS or by special scanningreceivers connected to the same antennas.

The same array has thus to receive, and to transmit, a multiplicity of signals for the MSwithin the cell coverage, and each antenna element shares all these signals. Theassociated beam forming network is complex and incorporates phase shifters, and - inthe case of adaptive arrays, amplitude control. Implementation of the BFN in the RFincurs a significant loss, that deteriorates the noise figure on reception, and theavailable power at the antenna for transmission.

Tx/Rx MULTI-BEAM ARRAY ARRANGEMENT

Fig. 2.8.: A multibeam ARM array

The use of ARM arrays revolutionize this application, as the active elements arelocated at the antenna terminals, the BFN losses are no longer of importance.Moreover, the BFN can be implemented at the baseband level, in digital processing, 10

and separate feed lines relayed to each element, or to each column (for horizontal-onlyscanning). This approach of feeding each column with its own power amplifier iscommon to most of the intelligent antennas proposed. However, these amplifiers areplaced at the BTS and each feeds a whole column - a similar situation to that discussedin Table 2.1. The ARM array offers a much higher reliability, due to their redundancy,in addition to all other advantages already mentioned. 2.5. Multiplex trunking of a multibeam / intelligent antenna

The number of cables laid on the tower, connecting the BTS and the antenna, is ofmajor concern, and affects the weight load, the cost and the complexity. The cablesassociated with ARM arrays are much thinner and lighter than otherwise, as they donot carry high power and may suffer a considerable loss with no degradation to thesystem. Nevertheless, the reduction of the number of these cables may be desirable incertain cases, and this option is offered by the multiplex trunking, described in Fig.2.9.This multiplexing, both on tx and Rx, is unique to ARM, where the trunking is notsensitive to losses

TxZRx MULTI-BEAM ARM ARRAY

BASE STATION TRANSMITTERS BASE STATION RECIEVERS

Fig. 2.9.: Tx and Rx multiplex trunking for multibeam/ adaptive arrays 11

2.6. ARM for CATV Infrastructure Based Remote Microcells.

The basic ARM is combined with a CATV up/down converter module to establish theCATV Infrastructure based Remote Microcells. This special application ARM makesuse of the existing CATV network as an RF trunk for remote RF Microcells. Such anarchitecture incorporates a great cost and capacity advantage. The block diagram ofthis ARM derivative is shown in Fig. 2.9.

Fig. 2.9.: ARM for CATV Infrastructure based Cellular - Block Diagram

The CATV up/down converter module input is a coaxial cable attached to the CATVnetwork and carrying the CATV standard forward link and reverse link. A bandwidthcommensurate with the cellular system has to be allocated in the CATV for eachforward and reverse links cellular ARM use.

The CATV diplexer within the converter separates the Tx and Rx signals. These arethen converted to the appropriate cellular frequencies. Each of these converters iscomposed of a mixer, Phased locked Oscillator and a Band Pass Filter to eliminateimage and L.O frequencies. The up/down converter module is attached directly to theARM in this application.

The ARM for CATV application will have a physical configuration suited for theapplication.

2.7. ARM Remote Microcell via FIBER A FIBER/RF converter module is attached to the basic ARM for FIBER-OPTICStrunking for remote ARM Microcells. RF trunking via fiber is an efficient method formicrocells layout, proposed for both in-buildings and outdoors Microcells distribution.The block diagram of this ARM derivative is shown in Fig. 2.10. 12

Fig. 2.10: ARM with Fiber-Optics conversion - Block Diagram

The Fiber/RF Transducer Module will contain both Tx Fiber/RF converter and RxRF/Fiber converter within the same module. The input to this module will be either onefiber carrying both Tx and Rx signals, split and filtered within the module, or twoseparate fibers, depending on Fiber infrastructure. The Fiber/RF converter module isattached directly to the ARM.

The same transducer unit will be used for high gain applications where desired. Thetransducers interface with the input/output of the beam forming network in a fixedbeam array, or with each column - for a multibeam/ adaptive array.

3. HIGH LEVEL DESIGN AND SPECIFICATIONS OF ARMAND ARM SUB-MODULES

The Active Radiator Module (ARM) is composed of six basic building blocks and theintegrating enclosure. Each one of these is described and separate specifications given. Thebasic building blocks are: 1. Tx amplifier 2. Rx amplifier 3. Tx/Rx Band Pass Filters 4. Tx/Rx antenna element 5. Power Supply 6. Monitor and Control (M&amp;C) circuit 7. Integrating Enclosure 3.1. Integrated ARM Specifications

The integrated ARM unit meets or exceeds the following specifications under fullduplex operating scheme and any combination of environmental conditions as specifiedherein. electrical• Tx Channel

Frequency band........................1960-1990 MHz (PCS)

Output power (average)............2 w for CDMA

Input power for max. output.....-2 dbm

Input power for bum out...............+10 dbm max. 13

Power control range.................20 db min

Power down at Shutdown........-50 db min

Output 1 db compression..........36-38 dbm 3ri order intercept point.............+46 dbm min.

CDMA ACP @+33dbm out.....-45dbc @1.25 MHz B.W

In/out Gain................................3510.5 db with compensation

Gain flatness.............................+/- O.ldb over any 1.25 MHz

Gain variation over temp............0.5 db max. with compensation

Transmission phase variation vs. Freq...........+/-1° over any 1.25 MHz

Transmission phase window between units ...+/- 5° AMZPM conversion...................0.25°/db Max up to 3 db below ldbcp

Noise figure...............................8 db Max

Spurious (non-harmonic)...........-60 dbc

Input VSWR............................1.5 : 1 @ 50 ohm system

Monitoring..............................see M&amp;C specifications

Antenna element Polarization....Vertical or Horizontal

Beam width @ -3db....................AZ 120 ° (-4 dB)

El 80°

Side lobes....................................El-T.B.Ddb

Front to Back ratio.....................TBD

Effective Radiated Power...........+38dbm • Rx Channel

Frequency range...........................1880-1910 MHz(PCS)

Noise figure.................................3.5 db max.

In/out Gain..................................30±0.5 db with compensation

Gain variation over temp..............0.5db max. with compensation

Input power for bum out.............+15 dbm max.

Power control range....................20 db min

Gain flatness...............................+/- 0. ldb over any 1.25 MHz

Transmission phase variation vs. Freq...........+/-1° over any 1.25 MHz

Transmission phase window between units...+/- 5°

Input 1 db compression point.......0 dbm min

Input 3rd order Intercept Point.....+10 dbm min

Spurious (non-harmonic).............-60 dbc

Class of operation........................A

Output VSWR............................1.5 : 1 @ 50 ohm system

Monitoring and Control...............see M&amp;C specifications

Antenna element Polarization....Vertical or HorizontalBeam width @-3db.....................AZ 120° (-4 dB)

El 80°

Side lobes....................................El -T.B.D db

Front to Back ratio.....................TBD • DC supply DC supply.................................+18v to +32v dc @

1.7A to 0.95A respectively per ARM DC connection...........................multiplexed on Tx input cable

Noise and ripple induced to input.. 10 mV peak max. up to 1 MHz

Lightning protection......................50v turn on,3 joules surge.

Thermal shut down.......................self contained within main DC/DC converter • Monitoring and Control

Monitoring and Control ............per following table 14

/Output Function Input Tx gain compensation and control Input Tx amplifier shut down Output Tx output power Output Tx input power Output Tx amplifier current sense Output ARM temperature sensor output Input Rx gain compensation and control Output Rx amplifier current sense M&amp;C communication with Base.......FSK modulated channel M&amp;C connection.............................multiplexed on Rx cable

Mechanical

Size...........................................140x70x160 mm max. weight.......................................1 lOOg max.

Radome.....................................fits outdoor use

Connectors...............................Tx and Rx connectors - N Type .female other types are optional.

Finish........................................White Polyurethane paint for outdoors.

Integration options.....................Mechanical structure enables integration into column or planar arrays.

Tilt option.................................mechanical tilt option up to -15° 3.2. Tx Amplifier A class AB amplifier with proper backoff and proprietary linearization serves therequirements for CDMA systems, and most other systems. An enhanced linearizationscheme will support all other multichannel-full bandwidth systems. Amplification isobtained in two stages: the first is a class A monolithic silicon, used also in PCShandsets. The second is a hybrid packaged power amplifier. The amplifier with itsmatching and biasing networks is assembled by SMT technology on a RF printedboard within the Tx amplifier enclosure.

Fig. 3.1 Tx Amplifier Block Diagram

Tx amplifier Specifications

Electrical

Frequency band........................1930-1990 MHz (PCS)

Output power (average)............2 w for CDMA 15

Input power for max. output.....-2 dbm

Input power for bum out...........+12 dbm max.

Power control range.................20 db min

Power down at Shutdown........-50 db min 1 db compression.......................36-38 dbm 3rd order intercept point.............+46 dbm min.

Two tones I.M products............-30 dbc for 1 w per tone -44 dbc for 0.2 w per tone

Gain..........................................35 db to 38 db, @ small signal without external compensation

Gain flatness.............................+/- 0. ldb over any 1.25 MHz

Gain variation over temp............3 db max. without external compensation

Transmission phase variation vs. Freq...........+/- 1° over any 1.25 MHz

Transmission phase window between units ...+/- 3° AM/PM conversion...................0.25°/db Max up to 3 db below ldbcp

Noise figure...............................8 db Max

Spurious (non-harmonic)...........-60 dbc

Input VSWR............................1.5 : 1 @ 50 ohm system output VSWR...........................1.3 : 1 @ 50 ohm system DC supply voltage.....................+8 volts and -5 volts DC with missing negative voltage protection DC supply current....................4Amp @8 v

Monitoring and Control............per following table funtfioft VI Input Tx gain compensation and control t.b.d Input Tx amplifier shut down t.b.d Output Tx output power t.b.d Output Tx input power t.b.d Output Tx amplifier current sense t.b.d Output ARM temperature sensor output t.b.d

Mechanical size...........................................t.b.d

Structure............................S.M.T P.C.B with trough holes for attachment 3.3. Rx Amplifier

The Rx amplifier within the ARM is connected to the Rx antenna output via the filter, andavoids the additional losses of the diplexer, beam forming network, and cable. Theredundancy of LNAs in the ARM array guarantees an extremely high reliability. The Rxamplifier has enough gain ,low enough noise figure, and high enough compression andintercept points to eliminate inter-channel interference and non-linear multi-channeldistortion, and degradation in sensitivity.

The communication in between the ARM and the base station is established by the M&amp;Csignal modulated on the Rx coaxial cable. The connection to the Rx cable is done by theRF/M&amp;C Diplexer at the output of the Rx amplifier , as shown in Fig. 3.2. The amplifiermeets the following draft specifications: 16

Rx AMPLIFIER BLOCK DIAGRAM

INDICATOR

TOP.S P.C.B TOM&amp;CP.C.B

Fig. 3.2. Rx Amplifier Block DiagramRx amplifier Specifications

Electrical

Frequency range......................................1850-1910 MHz (PCS)

Noise figure.............................................2.5 db max.

Gain.........................................................30db min. without external compensation

Gain flatness...........................................+/- O.ldb over any 1.25 MHz

Gain variation over temp..........................3 db max. without external compensation

Input power bum out..............................+15 dbm max.

Power control range...............................20 db min

Input 1 db compression point...................0 dbm min

Input 3rd order Intercept Point.................+10 dbm min

Transmission phase variation vs. Freq.....+/-1° over any 1.25 MHz

Transmission phase window between units ...+/- 3°

Spurious (non-harmonic).........................-60 dbc

Input/Output VSWR...............................1.5 : 1 @ 50 ohm system

Class of operation....................................A

Voltage supply........................................+8v regulated

Current requirement................................150 mA

Technology..............................................SMT of MMIC

Monitoring and Control...........................per following table input Input Rx gain compensation and control t.b.d Output Rx amplifier current sense t.b.d

Mechanical

Size.................................................................t.b.d

Structure...................................S.M.T P.C.B with through holes for attachment 17

3.4. Tx/Rx Band Pass Filters

Tx and Rx filters of the ARM establish .together with Tx/Rx antennas separation, adiplexer which isolates Tx and Rx signals from each other. Specifications for TX and Rxfilters are directly driven from performance requirements of ARM . 3.4.1. Tx Filter

The Tx filter has two roles within the ARM. The first is to reduce the Tx wide band noisein Rx band. The second is to reduce spurious signals which might interfere with Rxchannel of same cell or other base terminals, including other operators’ MS. The moredemanding requirement is the first one and it dictates the Tx filter performance and thusTx filter structure.

In order for the Tx noise that leaks into the Rx channel input to be lower or equal to theRx noise floor, isolation of 60 db (-74+135)is required. 20 db of the required isolation isattributed by Tx/Rx antenna isolation and the other 40 db plus 10 db of safety margin, aregiven by Rx band rejection of the Tx filter. 3.4.2 Rx Filter

The Rx filter has two roles within the ARM. The first is to reduce the Tx signal to a levelwhich does not interfere with the received signal causing Intermodulations , and thusDesensitization of the receiver channel. The other purpose of the Rx filter is to reduceinterfering signals from other Base Stations and mobile terminals . The more demandingrequirement is the first one and it dictates the RX filter performance and thus filter’sstructure.

In order for the Tx leakage not to interfere with received signal, it should be kept at amuch lower level than Rx channel compression for systems with no AGC or when AGC isat minimum. Assuming input ldb compression of -60 dbm at ARM input, the Tx leakageshould be below -70 dbm. For Tx average output power of +33dbm and Tx/Rx antennaisolation of 20 db, the Rx rejection of Tx band should be 75 db. A design goal for Rx filterrejection is 85 db to assure 10 db of margin. A dielectric filter at Rx amplifier input will beimplemented if the Rx filter rejection is insufficient. The same reasoning holds for CDMAsystems where numbers differ but ultimate result hold. 3.4.3. ARM Filters SpecificationsElectrical

Tx Filter

Pass band .................................... 1960-1990 MHz (PCS)

Rejection ..................................-50 db @50 MHz below pass band -40 db @ 50 MHz above pass band-60 db from 80 MHz above band to 4 GHz

Insertion loss................................-1.5 db max. @ pass band (0.5 db goal)

Ripple within band........................0.2 db max. over any 1.25 MHz band 0.6db max over 30 MHzGroup delay variation....................2 nsec max. over any 1.25 MHz

Transmission phase window between units...+/- 5°

Return loss in/out..........................-17 db min

Handling power.............................10 w max

Rx Filter

Pass band .................................... 1880-1910 MHz (PCS) 18

Rejection ..................................-75 db @50 MHz above pass band (-85 db design goal) -40 db @ 50 MHz below pass band-60 db from 80 MHz above band to 4 GHz

Insertion loss................................-1.5 db max. @ pass band (0.5 db goal)

Ripple within band........................0.2 db max. over any 1.25 MHz band 0.6db max over 30 MHz

Group delay variation.....................2 nsec max. over any 1.25 MHz

Transmission phase window between units ...+/- 5°

Return loss in/out..........................-17 db min

Handling power.............................10 w max

Fig 3.3. ARM Filters Performance

Mechanical Structure

Each of the Tx and Rx filters is a 6 coaxial resonators elliptic filter in combline structure.The housing is made of aluminum with tuning elements on the cover of filter’s housing asseen on Fig. 3.4.

Fig. 3.4. ARM Filters Mechanical Structure 19

3.5. ARM Antenna Elements

Both Tx and Rx antenna elements are printed elemental radiators. The Tx and Rx elementsare printed on same base material and covered by a radome of Epoxy-Fiberglass. Isolationof 20 dB between adjacent elements is achieved by the special design. The size of theARM front face - 0.44λ (H) and 0.88λ (V), is designed to allow full beam scanning andmultibeam arraying of ARM elements in the horizontal plane, and formation of a high gainarray in the vertical plane, as suitable for present and future applications in cellularsystems. The high isolation between the elements alleviates the need to retune the elementsupon arraying, and thus offers a simple engineering modular construction of the coverage,without resorting to additional R&amp;D and radiation testing.

The Antenna outline drawing is shown on Fig. 2.5.1. The antenna is environmentallyprotected by a Radome and is moisture sealed except for the interconnections.

Fig.3.5. ARM Antenna Outline Drawing3.5.1. ARM Antenna Specifications

Electrical

Frequency band...................Tx: 1960-1990 MHz for PCS

.................Rx: 1880-1910 MHz for PCS

Tx/Rx elements isolation.............20 db (15 db min.) for any Tx/Rx polarization combination

Polarization.................................Vertical or Horizontal in any combination

Gain...........................................5 dbi min. 6 dbi (target)

Beam width ...................AZ . 120. @-4db

El 80 ° max.

Side lobes....................................AZ: none

El: -15 db

Front to Back ratio.....................@90°-120° &amp; 240°-270°<-10 db @120°-240°<-15db 20

. 90% 1.6 : 1 Max

Efficiency................................... VSWR ,(@ 50Ω system).............

Mechanical

Size............................................

Connectors.................................

Radome......................................

Finish......................................... 140x70x15 mm

Coaxial connections per Fig. 2.4.2fits outdoor use ..White Polyurethane paint for outdoors. 3.6. Power supply

The ARM power supply has to supply all dc power requirement of the Tx and Rxamplifiers and include all protection means needed for a tower top mounted device,including Thermal over load protection and lightning secondary strike protection. SinceARM power supply is mounted on top of the antenna tower and the length of the cableconnecting the base station and the ARM is not fixed, a DC-DC converter is neededwithin this power supply. DC supply is done through the Tx coaxial cable which explainsthe BIAS-T is implemented within the Tx amplifier. The DC supply source is locatedwithin the base station. This way of DC supply is convenient for the modular approachwhere each module (CATV converter or Fiber/RF converter) has its independent powersupply, all consuming DC power from same source through connecting coaxial cables. ARM Power Supply Specifications

Electrical

Input voltage..............................18-32 v dc output voltages............................+8 v dc @ 4 Amp (or other voltage with same power) .............................-5 vdc@0.1 Amp

Regulation...................................± 2 %

Outputs Noise and ripple.............10 mV peak max. up to 1 MHz

Noise and ripple induced to input.. 10 mV peak max. up to 1 MHz

Lightning protection......................50v turn on,3 joules surge.

Thermal shut down.......................self contained with in main DC/DC converter

Mechanical

Size............................................T.B.D

Structure...................................S.M.T P.C.B with trough holes for attachment

Input connection...........................Through an internal BIAS-T 3.7 Monitoring and Control Circuit 3.7.1 General

The Monitoring and Control (M&amp;C) circuit controls the proper operation of the ARMcircuits and enables a real-time communication both ways between each individual ARMunit and the Base Station central computer. M&amp;C circuit is realized as a separate p.c.b ,integrated into the ARM assembly as part of the Rx subassembly. The interconnection inbetween Rx and Tx circuits is done by analog wiring. The dual directional communicationwith the base station is established through a FSK modulated communication channelmultiplexed on the Rx coaxial cable connecting the individual ARM ( or Rx BeamForming Network combiner of an array), to the base station. '21

The M&amp;C circuit tasks can be divided into two groups: Internal ARM functions andexternal ARM/Base station monitoring and control functions. a. ARM internal functions:

Tx amplifier gain compensation over temperature.

Rx amplifier gain compensation over temperature.

Thermal over load protection b. ARM/Base station external functions:

Individual ARM identification code.

Tx amplifier gain controlRx amplifier gain controlARM temperature sensingRx amplifier current sensingTx amplifier current sensingTx amplifier shut down 3.7.2 M&amp;C circuit specificationElectrical t.b.d

Lightning protection...............50v turn on,3 joules surge.

Mechanical

Size...................................T.B.D

Structure............................S.M.T P.C.B with trough holes for attachment

Input connection................Multiplexed on Rx cable through an internal BIAS-T at

Rx output. 3.8. ARM Integrating Enclosure

The integrating enclosure of ARM is based on the Tx/Rx filters block, which takes mostof the unit volume. Overall heat dissipated within the ARM is about 30 w, most of it onthe Tx block. This is dissipated by the die-cast Aluminum structure. Expected temperaturerise above ambient temperature is below 10 °c.

The antenna and radome are attached at the front of the ARM , while the Tx and Rxconnectors are located on the back side. The mechanical housing has the properarrangement for the mechanical interconnection of ARM units to form vertical and/orhorizontal arrays. A single ARM is easily replaceable on the mast in the arrayconfiguration. 22 3.9 Beam Forming Network

ARM Beam Forming Network Structure and Interconnections

Taw

N-TYTI

\COKN1CTOR

’ \ ItiOUT

\ \R-TYM

COMMCTOR

Fig 3.6 ARM B.F.N Structure and Inter-Connections ARM column Beam Forming Network

Fig. 3.7: 12 ARM Column Beam Forming Network DiagramBeam Forming Network Specifications

The following specifications are for 4 elements vertical arrays. Larger arrays are to beimplemented in the same basic structure:

Electrical

Frequency band.........................Tx: 1960-1990 MHz for PCS

.......................Rx: 1880-1910 MHz for PCS

Tx/Rx networks isolation..........60 db min.

Input to output insertion loss.....2 db max above division (summation) loss. VS WR at input/output. (50Ω system)...1.5 : 1 Max (1.3 goal)

Isolation in between ports.........20 db min.

Amplitude distribution...............equal on all outputs (or inputs)

Amplitude distribution error......0.75 db max. between outputs (or inputs)

Phase distribution......................equal on all outputs (or inputs)

Phase distribution error.............3°max. in between outputs (or inputs)

Permitted input power...............+23 dbm 23

DC connection..........................there should be a DC connection in between all outputs to the input of each B.F.N. DC current capability................up to 2 amp. at each Tx output

Mechanical

Size...........................................560x70x50 mm connectors................................to/from base: N Type, female to/from ARM: Floating OSP connectors

Finish........................................White Polyurethane paint for outdoors.

Sealing......................................The B.F.N is environmentally sealed 3.10 Weight of ARM module

Each ARM unit weight is less then 1100 gr.

Each High Gain ARM array of 4 units weight is less then 6 kgr.

Each High Gain ARM array of 12 units weight is less then 17 kgr. 3.11. Environmental Requirements

All ARM family modules and assembled structures shall exhibit in-spec, electrical andmechanical performance under all combinations of environmental conditions as listedhereinafter:

Environmental

Operating Temperature.............................-20 to +50°c

Non-operating temperature.......................-40 to +60°c

Operating altitude....................................8000 feet

Non-operating altitude.............................35,000 feet

Humidity.................................................up to 100% with condensation

Vibration.................................................5-50 Hz, sine, 0.1” displacement p.t.p

50-200 Hz, sine, 0.5 G

Wind load................................................200 km/h

Shock (non-operating).............................30G,half sine pulse, 11 msec

Salt atmosphere (non-operating)...............48 hours,5% salt solution per MIL-STD-202,method 101,condition b

EMC/EMI

Conducted emissions on cables.................T.B.D

Radiated emissions...................................T.B.D

Lightning Protection.................................±5kV input for 50 psec,1.2 μβεο rise and fall time

Electrostatic discharge............................±15kV surge by 500pf capacitor and series 150 Ω resistor

4. ARM ARRAY INTERFACE WITH THE BASE STATION

Interface description

Tx: The level of Tx power required for proper operation is 1 mW at the ARM input. A level of 5-10 mW is thus needed at the output of the BST which allows forlosses of the cables and BFN.

This may circumvent the High Power Amplifier rack at the BTS.

Rx ; The ARM LNA gain is 25db min. allowing for 5-10 db loss on the cable andBFN, the min. gain at the input to the BTS is 15 db.

This circumvents the LNA in the BTS 24

DC: The DC is provided via the Tx coaxial cable through an appropriate Bias-T

Supply voltage is 24-36 volts DC. Each ARM requires about 2 A.

Note: Remote High Gain Antennas with multiple ARM which require cable length over300 feet, may need a separate DC cable.

Monitoring and Control : The monitoring and control commands are transmitted over the Rx cable via an appropriate Bias-T. The data interface to hostcomputer is done by an Interfacing Control Box (ICB) using standard RS-232interface.

5. ARM-BASED BASE STATION - COST COMPARISON

The cost comparison is made between a cell fed by a 75W MCLPA and an 8 ARMsarray, and 2 -2 ARMs arrays in a transmit diversity

Table 5.1: BTS Sector RF chain cost budget ($) MCLPA Diplexer Cable -100’ Antenna ARM array Total Total 3sectors 2nd generation 20,000 1,000 350 500 21,850 65,550 ARM array 200 8,250 8,450 23,450 ARM Arrayswith Transmitdiversity 200 5,000 5,200 15,600

Note: the ARM array includes also the receive chain, not considered in the costcomparison. An additional cost difference of $3000.

This may be generalized by comparing the architectures of Fig. 1.1. and Fig. 1.2: a. The ERP $/Watt. b. The power delivered to the radiation $/Watt.

In a column of ARM elements, ERParm=N2Pa

With the MCLPA configuration (Fig 1.1.) ERPmclpa^NPmL where N is the number of elements in the array (and number of ARMs), L is the lossfrom the MCLPA to the radiating elements, and Pa, Pm are the power out from anARM unit and a MCPLA, respectively.

Now consider the cost Ca=KaN, Cm=KmPm

Then, by equating ERPs Cm/Ca= Km/Ka x Pa/L.

Here KA is the cost of an ARM unit, $/ARM, and KM is the cost coefficient of theMCLPA, $/Watt (assuming a linear proportion).

In this example KA= $1000, Km=$200/W, PA=2w, L=0.1 => Cm/Ca= 4

In an ARM column Pt=NPA, for the MCPLA P(=PmL

By equating the power delivered to radiation Cm/Ca= Km/Ka x Pa/L, which isthe same result.

The costs of the receive chain, and the additional elements in the transmit chain inFig. 1.1. configuration have not been incorporated in this calculation. 25

The ARM array offers additional advantages with high cost implications, that have tobe quantified for each market: • A smaller BTS cabinet and housing. The MCLPA occupies a significant portion ofthe BTS cabinets, and most of the power consumption and air-conditioningrequirements. The MCLPA is now eliminated, along with a substantial reduction inthe power and emergency power requirements, and in the air-conditioningrequirements. • There is no practical limit on the length of cables from the antenna mast to theBTS. The BTS may be located in an accessible and inexpensive location, notnecessarily very close to the antenna. • The overall weight on the mast is lighter. Though the ARM array is heavier thanthe passive antenna array, this is more than compensated by the lighter weightcables. • The reliability of the array of ARM elements is by far higher than that of a singleamplifier, and does not constitute an operation or a maintenance risk.

6. RELIABILITY ANALYSIS OF AN ARM ARRAY

6.1. Reliability of a single ARM

The system reliability is measured by MTBF - Mean Time Between Failures. Thisconsists of MTTF - Mean Time To Failure, and MTTR - Mean Time To Repair. Thefirst term is larger by far, and therefor MTBF=MTTF+MTTR=MTTF.

The formal evaluation of the MTTF of a system follows Mil-Std-756B. This elaborateprocess involves worst case environmental conditions, which render a very pessimisticvalue compared to the vast experience with similar commercial products, serving in theoutdoors. The latter will therefore be followed here, by quoting similar experience: • A 4 Watt VS AT outdoors RF head in C band which has been field proven. Suchunits have been produced and delivered by the same manufacturer, in quantities of2500 units per year, for the last 4 years. The rate of return for failures has been 2%per year, interpreted as an MTTF of450,000 hours. • The tracking antenna/RF/ down conversion head of the OmniTRACS, a Qualcommmobile USAT mounted on trucks, is known to have a similar rate of return. Thisunit is by far more complicated, and operates at Ku band. • On a component basis, a 4 Watt power amplifier device is known to have about 106hours MTTF. This is the critical element in the ARM.

Based on these, an evaluation by comparison is made that the MTTF of the ARM, inthe outdoors environment, is higher than 200,000 hours. 6.2. Reliability of an array of ARM elements

The array provides redundancy in the performance of each element, as shown inFig.6.1. _L__

Fig 6.1: An active antenna array 26

The ERP of the array relates to the total transmitted power x the gain of the array. Anarray with N ARMs, each transmitting P Watts, produces N*P Watts. The gain of acolumn array is also directly proportional to the number of the antenna elements N.The ERP of the ARM array is therefor proportional to N2P. A failure of a singleelement degrades the array ERP by less than 1 dB, as shown in Fig.6.2.

-Series 1-Series2-Series3-Series4-Series5

Fig. 6.2: deterioration of the ERP of a 8 element ARM array due to a failure ofone element a. Series 1 - the element pattern

b. Series 2 - the array ERP c. Series 3 - the array ERP, edge element missing d. Series 4 - the array ERP, element #2 missing e. Series 5 - the array ERP, element #4 missingA failure of x elements in a linear array: 1. Elements at the array edge. The remaining ERP is (n-x)2p, and the relative loss isΔ=(η-χ)2/η2. For n=10, x=l Δ=-.9 dB, x=2 Δ=-1.9άΒ 2. Elements not in the edge. In this case the gain is almost intact, except for raisingthe sidelobes. Δ=(η-χ)/η. For n=10,x=l, Δ= -.45 dB, x=2 Δ= -.96 dB. A failure of an element in a planar array, as shown in Fig. 6.3, is much less significant.

A failure of x elements in a planar array:

If the element is not on the edge Δ=1-χ/η2. For n=10, x=l Δ=-.Ο4 dB, x= 2 Δ=-.Ο8 dB.The values for elements on the edges are slightly higher. A failure of a single ARM in an array is not catastrophic, as shown above, and onlycauses a graceful degradation of the array performance. A failure of 2 elements ormore may be tolerated in an 8 element array before maintenance is called for. Thenumber of failures allowed in a planar array is much higher.

The failure probability is Pf=l/MTTF , and the reliability R=l-Pf=l-1/MTTF. TheMTTF of a failure of 2 ARMs in an array is derived by a conditional probability that asecond element fail, given one already failed. This is computed according to one of thealternative procedures(Mil-Std-756B): • Assume a model of “ N ARMs in series, in parallel to N-l ARMs in series”. R»=Rn ARMs+R(N-1)ARMs -Rn ARMsXR<N-1)ARMs

Rn ARMs-RaRM

Rt=l-N(N-l)PfARM2

When applied to Pf arm=1/200,000 , N=8 => Rt=l-1.4 x 10'9MTTF= 700,000,000 hours • According to“N-l units out of N must be working” model:

Rt=NxRARMN'1 - (N-1)RarmN= l-.5xN(N-l)Pf arm2When applied to Pf arm=1/200,000 , N=8 => Rt=l-.7 x 10'9 MTTF= 1,400,000,000 hours

Conclusion

The reliability of the array of ARM elements is by far higher than that of a singleamplifier, and does not constitute an operation or a maintenance risk. 28

PART TWO

SYSTEM APPLICATIONS WITH THE ARM CONFIGURATIONS

In addition to the performance improvement and cost saving, the ARM arrays offer ahost of new applications to the base station that further improve performance andcapacity. These arrays are also the essential building blocks for intelligent antennas.Among the applications are: • Polarization diversity and matching with ARM arrays. • Transmission (forward link) matching with ARM arrays. • Transmission diversity in CDMA with ARM arrays. • Multibeam and tracking antennas with ARM arrays. • Distributed microcell chain with ARM elements.

7.1. Polarization diversity and polarization matching with ARM 7.1.1. Background

The electromagnetic radiation is polarized, and allows for two orthogonal polarizationstates. Any antenna cannot be matched simultaneously to both polarization. Thepropagation of the signals through an non-homogeneous medium may transfer part ofit to the orthogonal polarization. This is the case for terrestrial communications, forexample, and in particular in urban areas, where the signals encounter multipaths fromobjects on the way. The transfer of polarization has been found to be typically -6 to -10dB in rural areas, and much more (-3 dB or higher) in urban areas and indoors. Theorthogonal polarization components have been found to have an independent fadingpattern, with correlation of less than .6.

Cellular systems operate traditionally in vertical polarization. Most of the mobileterminals, mounted on vehicles, have vertically polarized antennas. The polarization ofthe hand-held terminals is variable, however, and depends on the particular terminal, itsorientation vs. the head and the Base Station. There is an a-priori mismatch of thepolarization between the hand-held terminal and the BS antenna.

The forward (from the BS) and the reverse (from the MS) transmissions do not havecorrelated fading, because of the frequency difference between them. Diversityschemes, used for mitigating the fading, are therefore applicable on receive only, dueto lack of information to control transmission so as to reach the other receiver withoutfade. Only the BS are equipped with more than one antenna for receiving the signalfrom each MS, in most of the present day systems. The MS that are equipped with oneantenna only are unable to use antenna diversity for receive. The polarization matchingsystem described herein varies the polarization of the signal transmitted from the BaseStation so as to match the short-term average polarization of the MS. 29

7.1.2. The polarization diversity and matching A dual polarized antenna pair at the BS, with the appropriate receive channel for each,and the signal combining circuitry and control, adds polarization diversity to the BSreceiver, by itself or in addition to other diversities. The outputs of the diversitybranches reception are combined in an adaptive combining circuit at the Base Station.In the present polarization matching system the adaptive combining control circuit hastwo time constants: the fast circuit that adapts to the fading signals on the receivedreverse link, and changes the weights of the two receive branches, and a slow circuit,that follows only the physical movements of the MS, but averages the fading of thereceived signal - for each MS. This control is now applied to a dual polarizationtransmission ARM antenna, to form a polarization-matched transmission. This two-branch transmission is formed by splitting each transmit channel and weighing thesignals in proper phase and amplitude. A single RF channel transmission (e g. a single RF channel TDMA) can be weighted inRF, at the BA or by controlling the ARM amplifiers, for each time slot of the TDMAtransmission, as shown in Fig.7.1. Multi-channel transmission requires splitting eachsignal and weighting at the BTS, and summing up each branch before transmission tothe two ARM branches via the cables. This slow control, driven by the informationfrom the received signals, matches the transmitted signal polarization to that of theincoming signal, for each MS. This enhances the forward link significantly in situationsof large polarization mismatch, that may be frequent in normal operation, and are thecause of slow fading. 7.1.3. System description

H

Fig. 7.1: Polarization diversity/ matching circuits concept

Each of the receive antennas has its receive chain. The signal detected for each MS isweighed by weights a, then combined and controlled by the measurement and controlcircuit. This diversity combining is applied preferably at the baseband. These weightsare also fed into a transform circuit, that transforms the weights according to thedifferences in gain of the transmitting antennas. The low-pass filter averages the fastcontrol variations and responds only to the slow variations resulting from the physicalattitude changes of the MS. The transmit chain is split to two branches, and weightsare applied to each for each MS channel at the base band. The uniqueness of the ARM 30

architecture allows for the dual polarization transmit of the multi-channel signals,without the loss involved in the cables and combiners. Weights can be applied directlyto the ARM units in the case of single carrier TDMA systems (e.g. GSM minicells), bycontrolling the gain and phase. A simplified weighting scheme applies when the two branches operate at a fixedpower, which is the preferred case for the ARM: the phase of one of the branches isswitched between Οο,+9Οο,+18Ο°,-9Οθ,-18Οο. This sub-optimal weighting loses no morethan 3 dB in any state, as compared to the two branches polarization-matched.

The application details and system will be discussed upon request.

7.2. User tracking transmission from a base station using ARM

In this application a mobile communications system can enhance the transmitted signalto each mobile station(MS) while reducing the multipath fading, and the interference toother MS. 7.2.1. Background

The transmission in terrestrial environment encounters multipath due to scattering fromvarious objects. The multipath components, arriving at the receiver, interfere with eachother to form typical fading of the signal, when either transmitter, receiver or scatteringobstacles are in motion. This is one of the most detrimental effects on thecommunications in this service. Diversities are employed to mitigate the fading,whereby propagation routes which have an independent fading (non correlated fading)are received separately and then combined in a way that reduces the variations in thesignal. BS in the cellular service typically employs two receive antennas, spaced apartso as to have no correlated reception from the MS in their coverage area. Diversity onthe forward link (transmission from the BS) is not implemented in most systemsbecause of the excess complexity needed in the MS by introducing two antennas andthe accompanying circuitry. In transmission diversity from the BS, two or moretransmit antennas are positioned far apart to avoid correlation, and their transmittedsignals are controlled in amplitude and phase in correspondence to the output of theadaptive control circuitry of the receiving antennas. Such an arrangement does notapply to the cellular systems where the forward and the reverse links transmission havedifferent frequencies, far enough to decorrelate the fading between the two links.

Narrow beam transmission from the BS has been proposed, in order to increase thesignal strength received by the MS, and reduce the interference to other MS. Narrowbeam also reduces the multipath interference to the BS by limiting the illuminated areaaround the BS. Such a narrow beam has to track the desired MS direction. Algorithmsproposed for this task make use of the direction information obtained from thediversity antennas. A BS that communicates with many MS needs to formsimultaneously many narrow beams so as to optimize the transmission to each MS.Such an arrangement requires a multibeam antenna array - a relatively complex system,incorporating an antenna array with a complex beam forming network. Whilemultibeam arrays are most effectively implemented with ARMs, the use of two spacedcolumns of ARMS, as in space diversity, is a very cost-effective enhancement ofpresent base station performance. 31 7.2.3. Transmission gain with two antennas

Simultaneous transmission from two antennas forms a radiation pattern that ischaracterized by many radiation lobes, the width of each is inversely proportional tothe distance (in wavelengths) between the antennas. The amplitude of these lobes isbound by the radiation pattern of the individual antennas. Such a pattern, when aimedin a way to produce a maximum in the direction of a MS, has a gain of 3 dB in thatdirection. The narrow lobe around this maximum also reduces the scattering into theMS from the objects in the area that is now illuminated less. The typical signalsarriving at the MS from the scattering objects tend to diminish roughly in proportion tothe second power ofthe distance of these scatterers from the MS. Scatterers that areilluminated by the adjacent lobes are further away from the MS and contribute muchless to the fading at the MS. The contribution of the multilobe antenna to themitigation of fading, when its radiation is properly aimed, is significant and comesclose to that of a multibeam array whose size matches the distance between the twoantennas. This is illustrated in Fig. 7,2._

Fig. 7.2: Multipath suppression by the radiation lobes.

Series 1 - the multilobe radiation from the two antennas

Series 2 - the decay of the scatterers’ influence on the MS fading as a function of their distance from the MS.

Series 3 - the decay of the scatterers’ influence on the MS fading with the two-antenna transmission.

The interference to other MS is also reduced. Though some MS, positioned at thepeaks of other radiation lobes, receive the same signal strength as the desired MS,other MS within over 50% of the coverage area receive signals that are at least 3 dBlower, and over 10 dB lower within 20% of the area. 32

The application of such an arrangement is especially advantageous when the transmitantenna pair is positioned together with the diversity receive pair, on the sameinstallation. This is provided by a pair of ARM arrays. 7.2.4. System description

Two ARM arrays employed in this application, spaced apart as required for thediversity. The signals received from the two arrays are applied to the diversitycombining circuits at the BTS. The transmit antennas are weighed by a slow adaptivecircuit, responding to the short-term average of the received signal, after correctionsfor the difference in the frequencies of transmit and receive are made. These weightsdetermine the direction of the radiation lobes. There are many other directionestimation algorithms that can apply, however, and the choice depends on the systemand application.

Fig. 7.3: The multilobe transmission by the antenna pair

The application details and system will be discussed upon request. 33

7.3. Transmission diversity from the CDMA Base Station with ARM

7.3.1, Background

The forward and the reverse links in most of the cellular systems are transmitted atdifferent frequencies (what is called FDD - frequency division duplex). The typicalspacing for the 800 or 900 MHz systems is 45 MHz, which is larger than the coherencebandwidth of the typical terrestrial cellular communications channel. There is notenough correlation between the fading in both channels, and the fading informationfrom the reverse link cannot be used to control a diversity transmission in the forwardlink. A unique feature of the CDMA system is the “rake receiver”, which consists of anumber of digital correlators that enable the reception of different transmissions, and -in particular, the same transmission with a different delay. The receivers in the systemIS 95, for example, are equipped with 4 such correlators (called “fingers”) that, in theintended mode of operation, search for delayed multipath transmission. This featurecan be used for intentionally delayed transmission from two or more antennas whosefading are not correlated, e.g. antennas that are spaced apart far enough, or transmitwith different polarization. 7.3.2. System description

The transmission signal is split to the number of ARM antennas, as in Fig. 7.3..Typically two antennas. A delay that is longer than the CDMA chip, but preferablyabout 5 gs in this case, is inserted in the transmit chain of each of the other antennasrelative to the first antenna. This delay may be inserted in the RF chain, by a delay line,or in the digital base band. The signal is then transmitted by all antennas, each havingabout the same coverage. These signals are then received by the MS with different“fingers” and optimally combined, as with natural multipath. The diversity gain thusobtained may be significant, as it is for the reverse link.

The typical delay profile depends on the environment. In urban, and suburbanenvironments the first delay cluster may extend to about 2 microseconds. An optimaldelay has to be selected for the transmission, in order to minimize the interference from natural multipath arriving at the same time.

Fig. 7.3: Transmit antennas for space diversity

The transmit diversity is implemented with two ARM columns as in Fig. 7.3. A singletransmit low power RF cable is required between the BS and the antenna pair. The 34

signal is split at the antenna back plate module, and one branch is then delayed by aSAW delay line, and then amplified to an equal level to that of the other branch. Thisextremely simple arrangement is unique to the ARM, and is described in section 2.3and in Fig.2.6. In an alternative arrangement the signal is split, and then transformed tofiber, where one branch is then delayed by a length of fiber. While this latter may bemore expensive, it may be the preferred solution in case where it is advantageous torun the fiber to the antenna instead of RF cable.

The application details and system will be discussed upon request. 7.4. Multibeam reception and transmission in CDMA IS95 7.4.1. Background

Narrow beams have the advantages of higher gain and of reduced interference, andthereby have the potential to benefit the cellular base station. The implementation ofnarrow beams requires the mechanism for aligning the subscriber receivers to thebeams with optimal reception of these subscribers. The other difficulty is inimplementing receive diversity, which may require two arrays, properly spaced, orbeam diversity.

The receiver of Base station of the IS 95 system is presently built to incorporate up to6 receive antennas, for diversity reception in 3 sectors. Each receiver output feeds adigital sample bus, that is sampled by each channel card. In this way, each channel cardscans 4 to 6 antenna inputs for the best signal. With the multibeam arrangement of Fig. 2.8, or Fig. 2.9. (with multiplex trunking), (sec. 2.4,2.5) the outputs of the beams, astransformed by a Butler matrix or equivalent transform circuit, are connected to theantennas inputs in the CDMA BS. This way the whole BS is dedicated to the coveragescanned by the multibeam array. The BS automatically searches for the right beam (orbeams, for diversity reception) for each channel. The capacity of the reverse linkincreases by the number of the beams in this arrangement, while using the existing BShardware.

The transmit arrangement for CDMA does not necessarily have to follow the samebeam shapes. In case that the forward link is much stronger than the reverse one, morethan one receive beam can be accommodated in a single transmit beam (or sector). Thetransmit beam can be shaped to follow coverage requirements, and be enhanced bytransmission diversity. The latter is done according to the discussion in sec. 7.3., bysplitting the transmit array into two halves. The beam shaping is done by weighting thebeam inputs to the required shape in Fig. 2.8 or 2.9.

Full transmit multibeam in CDMA is also possible, and may multiply the capacity N-fold (N being the number of beams). This requires a change in the BS digital rack,however; beam interface cards have to replace the sector interface cards, and eachchannel be assigned to the proper beam for summation and transmission. The beamsthus replace the function of the sectors, and the BS has to have more of them. Theassignment of the transmit channel to the proper beam should follow the receivechannel beam selection, with proper low-pass filtering as in sec. 7.1,7.2.

Note: US Patent # 5,563,610, awarded in 1997, claims for a similar, yet differentarrangement. 35 7.5. Distributed microcells chain with ARM elements * 7.5.1. Background

High density areas, and urban areas that require good coverage of the low levels inpublic buildings, may prefer a microcell distribution over high-positioned macrocells.The compactness, cost ands simplicity of the distribution system are key parameters indeploying microcells. There is an appreciable advantage to a chain of RF microcells,whereby the RF unit only is deployed, and back-connected to the BS by RF, fiber orCATV. The RF network allows for certain spatial processing that is impossiblebetween cells, e.g. diversity. 7.5.2. ARM chain of microcells

The ARM is a self contained RF unit. Its power output is sufficient for single channelCDMA, and for comparable GSM or IS 136. A two-ARM package can provideenhanced power and diversity wherever required. The antennas may be replaced byomnidirectional antennas. A single ARM is sufficient for a single channel CDMAmicrocell, and comparable GSM or IS 136. The RF linkage to the central cell allowsfor diversity and other spatial processing schemes, avoiding the handoff process,impossible with full microcells. The gain monitoring and control allow for balancingthe network. ARM microcells, trunked by RF, CATV, or fiber, are described insec.2.2,2.6,2.7. A distributed antenna for CDMA is described in sec.2.3.

7.6. ARM cell - guidelines for cell design with ARM

An important issue in the design of a cell is the design of the transmitted power, andantenna gain, as this determines the coverage and the capacity of the cell/ sector. A cellI sector based on ARM array (for shortage - ARMcell) has a linkage between thenumber of ARM units, their arrangement and the cell coverage and capacity. 7.6.1. Link balancing

The link budget is

y = GtGrL s T where Pr= Power at the receiver antenna outputPt=Power at the transmit antenna inputGt,Gr gain of the tx and Rx antennas, respectivelyL path loss T transmission loss

Now, the minimum receive power for a given service in

PrM=SNR(signal to noise ratio) x N(noise) x NF(noise figure of the receiver)

and Pim=SNR x N x NF / T

The transmit power required from the base station can thus be inferred from the powertransmitter by a MS(Mobile Station) times the number of MS - n - that the BS serves,by comparing the SNR, T and NF of the forward and reverse links: 36 P =

1 t.BS

T^SNR.N.NF)^TK)(SNR.N»NF),s ‘ 7.6.2. ERP and gain with ARM columns

The gain of a linear antenna array is roughly λ λ where D/λ is the total length (height) of the antenna array, in wavelengthsN is the number of ARM elements d/λ is the distance between two adjacent ARM units in wavelengthsS is the numbers of sectors.

For an ARM linear array it becomesa®] = 7.25 + 1 OZogV

The ERP of a linear ARM array is = G ' 3NPm = =532^+^ =10.64JV: and in dBw ERP[Watis] = 10.27 + 20Log(N) 7.6.3. CDMA IS 95

The forward link is coherent, pilot-aided and orthogonal (partially), and the required SNR (or ) is 2 -3 dB lower than that of the reverse link (thins already includesthe effect of antenna diversity on the reverse link). On the other hand - the NoiseFigure of the MS receiver is 3 dB higher than that of the BS, and altogether theyequalize. Therefor, if the antenna gains for both links are the same

P ~nP

rt,BS — rtrt,MS

Note that this result does not depend on the range of the cell. This enters through theBS antenna gain ( a higher gain suppresses T) and through Pt,Ms· Higher power may berequired from the MS on the margin of range-limited cells than in capacity-limited cellsand microcells.

Microcell t,Ms <50 mWatt in the microcell, and each sector serves 20 MS per channel, thenPt,Bs < 1 Watt which is the less than the power provided by a single ARM unit (2 Watts).

Large cell

For a loaded large cell, single channel, Pi,ms < 200 m Watt, and Pi,bs < 4 Watt.

This is served by 4 ARM units. When stacked into a column they form an array thatalso provides about 11 dB gain. More ARM units may be stacked either in a column,to provide further gain, or in two shorter columns. Note that this is an extremeassumption, and pi,ms < 100 mWatt is more realistic.

Forward-link diversity CDMA has the unique capability of providing forward link diversity from the basestation by transmitting from two displaced antennas and providing the appropriate 37

delay between them. This is implemented in a simplest way by the ARMcell, by twospaced ARM units or columns. The diversity gain provided, nominally higher than 3dB, may be used to reduce the required P(,bs with a considerable cost saving to thebase station. Thus a 2 x 4 ARM cell/ sector can accommodate 16 RF channels, fullyloaded, when the required MS power does not exceed 100 mWatt. With an averageload of 50% this ARMcell accommodates the full 30 MHz allocation.

7.6.4. AMPS / IS 136 / GSM

The forward link is weaker in these systems than the reverse link: as much as themodulation is the same, the reverse link enjoys both the antenna diversity gain (> 3 dB)and the lower Noise Figure of the BS receiver ( 3 dB), and the reverse link is strongerby as much as 6 dB. This accounts for the higher transmit power required from thebase stations in these technologies.

Microcell

If Pt,Ms < 100 mWatt, a single ARM can support 2 RF channels (2 AMPS or 6 IS 136or 15 GSM MS). Most microcells do not support more than 2 RF channels in presentconfiguration.

Large cells

The highest power class of MS transmits 2 Watts. It is reasonable to assume that theaverage will not exceed 500 mWatt. One ARM is then required for each RF channel ina large cell. A column of 12 ARM units serves 12 RF channels (36 IS 136 or 95 GSMMS), creating a gain of about 17 dB. An alternative arrangement of 2 x 8 elementsprovide 16 Rf channels (48 IS 136 or 127 GSM MS), with a gain of 15 dB.

Transmission matching of direction or polarization in the forward link

These applications are simply implemented in ARMcells, and gain about 3 dB to theforward link, thus halving the required transmit power, and the number of therespective ARM units. 38

Contents33

28 members in 8 offices

This recordMember, by publication datePriority claim

Priority claims2

Priority claims
Document	Office	Kind	Date
12120197	Israel	A	1997-06-30
IL19970121201	–	–	–

Members28

Family members, oldest first
Document	Office	Kind
WO9839851A1	World Intellectual Property Organization (WIPO)	A1	1998-09-11
WO9839856A1	World Intellectual Property Organization (WIPO)	A1	1998-09-11
AU6113198A	Australia	A	1998-09-22
AU6228898A	Australia	A	1998-09-22
EP0916195A1	European Patent Office (EPO)	A1	1999-05-19
EP0916195A4	European Patent Office (EPO)	A4	2000-03-08
EP1012994A1	European Patent Office (EPO)	A1	2000-06-28
JP2000509950A	Japan	A	2000-08-02
KR20000065190A	Republic of Korea	A	2000-11-06
IL120364A	Israel	A	2000-11-21
IL120706A	Israel	A	2000-12-06
IL121201AThis record	Israel	A	2000-12-06
KR20000075941A	Republic of Korea	A	2000-12-26
WO0106595A2	World Intellectual Property Organization (WIPO)	A2	2001-01-25
AU6010000A	Australia	A	2001-02-05
JP2001513969A	Japan	A	2001-09-04
WO0106595A3	World Intellectual Property Organization (WIPO)	A3	2001-11-22
EP1203421A2	European Patent Office (EPO)	A2	2002-05-08
US2003073463A1	United States of America	A1	2003-04-17
US6640110B1	United States of America	B1	2003-10-28
US6640111B1	United States of America	B1	2003-10-28
US6697641B1	United States of America	B1	2004-02-24
CN1579035A	China	A	2005-02-09
US2005075139A1	United States of America	A1	2005-04-07
US6900775B2	United States of America	B2	2005-05-31
KR20050098028A	Republic of Korea	A	2005-10-10
KR100521854B1	Republic of Korea	B1	2005-10-14
US7072611B2	United States of America	B2	2006-07-04

5 legal events, as the office reported them to INPADOC

Over the term

Point at a mark for the event

Events

Legal events, newest first
	Event	Code
2012-02-29	Patent not in force due to non-payment of renewal feesMM9K	MM9K
2007-08-19	Patent renewedKB	KB
2004-02-08	Patent renewedKB	KB
2001-07-24	Patent renewedKB	KB
2001-04-30	Patent grantedGrantedFF	FF

Abstract

Term

CELLULAR COMMUNICATIONS SYSTEMS

CELLULAR COMMUNICATIONS SYSTEMSFIELD OF THE INVENTION

BACKGROUND OF THE INVENTION

SUMMARY OF THE INVENTION

BRIEF DESCRIPTION OF THE DRAWINGS

DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS

APPENDIX A

CONTENTS

CORPORATEFEEDDIVIDER

A. R. M

TO MODIFIEDBASE STATION

ARM

SECTORIAL

ANTENNA

ARRAY

CORPORATE

FEED

FU

A. R. M

A. R. M

BASE STATION TRANSMITTERS BASE STATION RECIEVERS

3. HIGH LEVEL DESIGN AND SPECIFICATIONS OF ARMAND ARM SUB-MODULES

INDICATOR

N-TYTI

COMMCTOR

EMC/EMI

4. ARM ARRAY INTERFACE WITH THE BASE STATION

5. ARM-BASED BASE STATION - COST COMPARISON

6. RELIABILITY ANALYSIS OF AN ARM ARRAY

PART TWO

SYSTEM APPLICATIONS WITH THE ARM CONFIGURATIONS

H

7.6.4. AMPS / IS 136 / GSM

Priority claims2

Members28

Over the term

Events

Numbers

Titles

Classification