Spectrally efficient high capacity wireless communication system with spatio-temporal processing

Spectrum high-efficiency high-capacity wireless communication system with time-space processing

This application is the continuation of the pending US Patent Application Serial No. 08/375,848 filed on January 20, 1995, entitled "Spectrum High Efficiency and High Capacity Wireless Communication System," and it is also filed on December 12, 1991 under the title US Patent Application Serial No. 07/806,695 for "Space Division Multiple Access Wireless Communication System" and US Patent Application Serial No. 08/234,747 entitled "Method and Apparatus for Calibrating Antenna Array" filed on April 28, 1994 The continuation part.

Wireless users use remote terminals such as cellular phones and data modems equipped with wireless transceivers to access wireless communication systems. The system (especially the remote terminal) has protocols for initial calls, receiving calls, and general information transfer. Information can be transferred in real time in the case of circuit-switched voice calls and faxes, for example, or in a store-and-forward manner in the case of e-mail, paging and other similar messaging systems, for example.

Usually a part of the radio frequency spectrum is allocated to wireless communication systems for their work. Divide the allocated frequency spectrum into multiple communication channels. These channels can be distinguished by frequency, time, code, or some combination of the above. Each of these communication channels is referred to herein as a channel. In conventional communication systems, these are designed as separate or non-overlapping (in time, frequency, and/or code) channels, which are referred to herein as conventional channels. Here, these channels share common resources, and they can be non-overlapping, partially overlapping, or completely overlapping. The wireless system can have from one to hundreds of communication channels depending on the available frequency allocation. In order to provide a full-duplex communication link, some communication channels are usually used for communication from the base station to the user's remote terminal (downlink), and others are used for communication from the user's remote terminal to the base station (uplink).

A wireless communication system generally has one or more wireless base stations, each of which provides coverage to an area called a cellular, and is often used as a point-of-presence (PoP) to provide services such as the Public Switched Telephone Network (PSTN) ) And other WAN connections. Each wireless base station is often assigned a predetermined subset of available communication channels in an attempt to minimize the amount of interference experienced by system users. In its cell, the wireless base station can use the different conventional communication channels of each remote terminal to communicate with many remote terminals at the same time.

As mentioned above, the base station can function as a PoP, providing connections to one or more wired communication systems. These systems include local area data networks, wide area data networks, and PSTN. Therefore, remote users can be provided with access to local and/or wide-area data services and local public telephone systems. Base stations can also be used to provide local connections without direct access to wired networks such as local emergency and mobile theater communication systems. The base station can also provide various connections. In the above example, it is assumed that approximately the same amount of information flows in the two directions between two users in peer-to-peer communication. In other applications such as interactive television, the base station broadcasts information to all users at the same time, and processes responses from many remote units.

However, the spectrum efficiency of conventional wireless communication systems is relatively low. In a conventional wireless communication system, only one remote terminal can use any conventional channel in a cell at any time. If more than one remote terminal in a cell tries to use the same channel at the same time, the downlink and uplink signals associated with the remote terminal interfere with each other. Since conventional receiver technology cannot eliminate interference in these combined uplink and downlink signals, when interference occurs, the remote terminal cannot effectively communicate with the base station. Therefore, the total capacity of the system is limited by the number of regular channels that the base station can provide, and in the entire system, is limited by the way these channels are reused in multiple cells. Therefore, conventional wireless systems cannot provide capacity near wired communication systems.

In the pending U.S. Patent Application No. 08/375848 filed on January 20, 1995, entitled "Spectrum High Efficiency and High Capacity Wireless Communication System", we have disclosed the use of antenna arrays and signal processing to separate combined reception ( Uplink) signal. We also disclose the use of transmit spatial multiplexing for downlink signals. As a result, the spectrum efficiency, capacity, signal quality, and coverage area of the wireless communication system are improved. By allowing multiple users to share the same communication channel in a cell at the same time without interfering with each other, and further by allowing more frequent reuse of the same channel in an area that covers many cells, the capacity is increased. Improve signal quality and coverage area by appropriately processing signals received from and transmitted by multiple antenna elements. In addition, the purpose of the original application No. 08/375,848 and the invention described herein is to provide capacity increase by dynamically allocating channels among multiple base stations and remote terminals.

In short, the invention of the original application No.08/375,848 includes an antenna array and a signal processing device for measuring, calculating, storing, and using the spatial signature of the receiver and transmitter in the wireless communication system to improve the system Capacity, signal quality, coverage, and reduce the cost of the entire system. Antenna arrays and signal processing devices can be used in base stations (PoP) and remote terminals. Generally speaking, base stations in many signal concentrations may have different processing requirements than remote terminals that usually manage only a limited number of communication links.

As an example, in wireless local loop applications, a particular base station can serve as a PoP for many remote terminals and use the antenna arrays and signal processing devices described here. In addition, the remote terminal can use antenna arrays and signal processing devices to further improve its capacity and signal quality on a simpler remote terminal that handles several communication links. Here, the difference between a base station and a remote terminal is that the base station usually functions as a concentrator that connects to multiple remote units at the same time, and can provide a high-capacity connection to a wide area network. For the sake of clarity, the following discussion is presented here on the basis of a simple remote terminal that does not use an antenna array, which should not be construed as limiting the application here. Therefore, the spatial signature will be mainly related to the remote terminal from now on. When the antenna array is used in the remote terminal, the base station will also have the related spatial signature.

In short, as described in the original application No.08/375,848, there are two spatial signatures associated with each remote terminal/base station pair on a specific frequency channel. For the purpose of this discussion, it is assumed that only the base station With antenna array. The base station passes the spatial signature related to how the remote terminal receives the signal transmitted to it by the antenna array of the base station, and the second spatial signature related to how the receiving antenna array of the base station receives the signal transmitted by the remote terminal. Related to a remote terminal. In a system with many channels, each remote terminal/base station pair has a transmit and receive spatial signature for each channel.

The receiving spatial signature is characterized by how the base station antenna array receives signals from the specific remote unit in a specific channel. In one embodiment, it is a complex vector including the response (amplitude and phase relative to the reference) of each antenna element receiver, that is, for an m-element array, hr=[hr1,hr2,...,hrm] T,--(1)]]>where It is the response of the i-th receiver to the unit power transmission signal from the remote terminal. Assuming that the narrowband signal sr(t) transmitted from the remote terminal is transmitted from the remote terminal, the output of the base station receiver at time t is given by the following formula: xr(t)=hrsr(t-τ)+nr(t), ( 2) Where τ indicates the average propagation delay between the remote terminal and the antenna array of the base station, and nr(t) indicates the noise that appears in the environment and the receiver.

In the original application No. 08/375,848, the transmitted spatial signature is characterized by how the remote terminal receives signals from each antenna array element of the base station in a specific channel. In one embodiment, it is a complex vector containing the relative amount (amplitude and phase relative to the reference) of each antenna element transmitter output included in the remote terminal receiver output, that is, for an m-element array, ht=[ht1 ,ht2,...,htm]T,---(3)]]>where It is the amplitude and phase (relative to a fixed reference) of the remote terminal receiver output relative to the unit power signal transmitted from the i-th element in the base station array. Assuming that the vector st(t)=[st1(t),...,stm(t)]T of the complex signal is transmitted from the antenna array, the output of the remote terminal receiver is given by the following formula: zt(t)= htTst(t-τ)+nt(t),---(4)]]> where nt(t) represents the noise in the environment and the receiver. These spatial signatures of each remote terminal and each channel in its cell are calculated (estimated) and stored at each base station. For fixed remote terminals and base stations in a stable environment, the spatial signature may not be updated frequently. However, changes between the base station and the remote terminal in the RF propagation environment can often change the signatures and require them to be updated. It is pointed out later that the time argument in parentheses is suppressed; only integers are used to index vectors and matrices in parentheses.

In the above discussion, it is assumed that the receiver and transmitter time match. If there are differences in the time response, the well-known temporal filtering techniques can be used to equalize these differences. In addition, it is assumed that the channel bandwidth is smaller than the center frequency of operation. Large bandwidth channels may require more than one complex vector in order to accurately describe the well-known output.

In the original application No. 08/375,848, when more than one remote terminal wants to communicate at the same time, the signal processing device of the base station uses the spatial signature of the remote terminal to determine whether a subset of them can communicate with the base station at the same time through the shared channel. In a system with m receiving and m transmitting antenna elements, up to m remote terminals can share the same channel at the same time.

When multiple remote terminals share a single uplink channel, the multiple antenna elements of the base station each measure the combination of the arriving uplink signal and noise. These combinations result from the relative position of the antenna elements, the position of the remote terminal, and the environment of RF propagation. The signal processing device calculates the spatial demultiplexing weight to separate the uplink signal from the combination of the uplink signal measured by the multiple antenna elements.

In the application of transmitting different downlink signals from the base station to the remote terminal, the signal processing device calculates the spatial multiplexing weight used to generate the multiplexed downlink signal. When transmitting from the antenna element of the base station, the spatial multiple Multiplexing weighting results in the reception of the corrected downlink signal with appropriate signal quality at each remote terminal.

In the case of transmitting the same signal from the base station to a large number of remote terminals (greater than the number of antenna elements), the signal processing device calculates a weight suitable for the broadcast signal to cover the area required to reach all remote terminals.

Therefore, in the original application No. 08/375,848, the signal processing device facilitates simultaneous communication between a base station and multiple remote terminals on the same channel. The channel can be a frequency channel, a time slot in a time division multiplexing system, a code in a code division multiplexing system, or any combination of the above. In one embodiment, all elements of a single antenna array transmit and receive radio frequency signals, while in another embodiment the antenna array includes separate transmit antenna elements and receive antenna elements. The number of transmitting and receiving antenna elements need not be the same.

When there is a wideband channel and/or there is significant delay spread or scattering, the well-known time equalization is used. In the original application No. 08/375,848, if necessary, it is assumed that the time equalization is performed after spatial demultiplexing. The channelization in the FDMA (or CDMA) system is to perform filtering in order to separate out the frequency (or coding) channel, and perform it before the spatial processing. This decoupling of spatial processing from temporal processing such as equalization and channelization is probably not optimal, and combining spatial and temporal processing may have performance advantages. Therefore, there is a need in the art for methods and devices that define space and time processing together as a single spatio-temporal processing step, as well as methods and devices for performing the spatio-temporal processing.

In short, the present invention includes an antenna array and a signal processor for measuring, calculating, storing and using spatiotemporal signatures of receivers and transmitters in wireless communication systems to improve system capacity, signal quality, and coverage. , And reduce the cost of the entire system. The antenna array and signal processor can be used in base stations (PoP) and remote terminals. Generally speaking, base stations in many signal concentrations may have different processing requirements than remote terminals that usually manage only a limited number of communication links.

As an example, in wireless local loop applications, a particular base station can serve as a PoP for many remote terminals and use the antenna arrays and signal processing devices described here. In addition, the remote terminal can use antenna arrays and signal processing devices to further improve its capacity and signal quality on a simpler remote terminal that handles several communication links. As the term used here, the difference between a base station and a remote terminal is that a base station usually functions as a concentrator that connects to multiple remote units at the same time, and can provide a high-capacity connection to a wide area network. For simplicity, most of the descriptions here are systems with simple remote terminals that do not use antenna arrays. However, it should not be interpreted as limiting the application. Therefore, from now on, the spatio-temporal signature will be mainly related to the remote terminal. When the antenna array is used in the remote terminal, the base station will also have the relevant spatio-temporal signature.

Each base station calculates (estimates) and stores the spatiotemporal signature of each remote terminal and each channel in its cell. For fixed remote terminals and base stations in a stable environment, the spatio-temporal signature may not be updated frequently. However, changes between the base station and the remote terminal in the RF propagation environment can often change the signatures and require them to be updated.

When more than one remote terminal wants to communicate at the same time, the signal processor of the base station uses the spatio-temporal signature of the remote terminal to determine whether a subset of them can communicate with the base station at the same time through the shared channel.

When multiple remote terminals are using overlapping uplink channels, the multiple antenna elements of the base station each measure the combination of the arriving uplink signal and noise. These combinations result from the relative position of the antenna elements, the position of the remote terminal, the frequency characteristics of the receiver and transmitter, the spectral content of the signal, and the RF propagation environment. The signal processor calculates the space-time demultiplexing weights to separate the uplink signal from the combination of the uplink signal measured by the multiple antenna elements.

In the application of sending different downlink signals from the base station to the remote terminal, the signal processor calculates the weights of space-time multiplexing used to generate multiplexed downlink signals, and transmits the space-time multiplexing from the antenna elements of the base station The downlink signal results in the reception of a corrected downlink signal with appropriate signal quality at each remote terminal.

In the case of transmitting the same signal from the base station to a large number of remote terminals, the signal processor calculates the space-time transmission weight suitable for the broadcast signal to cover the area required to reach all remote terminals.

Therefore, the signal processor facilitates simultaneous communication between a base station and multiple remote terminals on overlapping channels. The channel can be a frequency channel (Frequency Division Multiple Access, FDMA), a time slot in a Time Division Multiplexing system (Time Division Multiple Access, TDMA), an encoding in a Code Division Multiplexing system (Code Division Multiple Access, CDMA), or any combination of the above. The channel can also be composed of multiple regular channels.

In one embodiment, all elements of a single antenna array transmit and receive radio frequency signals, while in another embodiment the antenna array includes separate transmit antenna elements and receive antenna elements. The number of transmitting and receiving antenna elements need not be the same.

The present invention and its purpose and characteristics will become more apparent from the following description in conjunction with the accompanying drawings and appended claims.

There are two spatiotemporal signatures associated with each remote terminal/base station pair on a specific frequency channel. For the purpose of discussion, it is assumed that only the base station has an antenna array. The base station uses the transmission spatio-temporal signature related to how the remote terminal receives the signal transmitted to it by the antenna array of the base station, and the reception spatiotemporal signature related to how the receiving antenna array of the base station receives the signal transmitted by the remote terminal. Related to a remote terminal. In a system with multiple channels, each remote terminal/base station pair has a transmit and receive spatiotemporal signature for each channel.

The receiving spatio-temporal signature is characterized by how the base station antenna array receives signals from the specific remote unit in a specific channel. In one embodiment, it is a matrix including the impulse response of the antenna element receiver as described below.

Assume that the signal sr(t) is transmitted from a remote terminal. Let m be the number of antennas and associated receivers in the base station. Then in an embodiment, the output of m base station receivers at time t can be expressed as xr(t)=xr1(t)xr2(t)...xrm(t)=hrsrMr(t-τ)+nr (t),---(5)))> where srMr(t-τ)=sr(t-τ)sr(tT-τ)...sr(t-(Mr-1)T- τ)---(6)]]>hr is the channel response matrix. In this embodiment, it is assumed that the finite impulse response filter is the accurate feature. Let τ denote the average propagation delay between the remote terminal and the base station antenna array, and T is the sampling time, and in this embodiment it is assumed that it satisfies the well-known Nyquist sampling theorem. Mr is the channel response length, nr(t) represents the noise in the environment and the receiver. The channel response of the receiver i at the antenna element is given by the row vector hr(i). The channel response matrix is a collection of individual channel responses hr=hr(1)...hr(m)---(7)]]>If the impulse response of the communication channel, antenna element, receiver and transmitter filter The length of is Mr, and the impulse response is equal to the channel response matrix hr. If the impulse response lasts longer, the channel response matrix hr is an approximation of the impulse response, and the generated error is added to the noise term nr(t). In one embodiment, the output of the delay receiver, referred to herein as the space-time reception vector zr(t), is established by the following formula: zr(t)=xr(t)xr(tT)...xr(tT(Lr- 1))=HrsrMr+Lr-1(t-τ)+er(t).---(8)))> where Lr is the length of the sliding windower(t)=nr(t)nr(tT)...nr(tT(Lr-1))---(9)])>mLr×(Mr+Lr-1) The matrix Hr is called in a On a specific channel, the remote terminal transmits sr(t) and the base station receives the received spatio-temporal signature of zr(t).

When multiple remote terminals are active on the same channel, a single received spatio-temporal signature is collected in the demultiplexed spatio-temporal signature matrix Hr. For each channel, Hr is formed by a single spatiotemporal reception signature: Hr=[Hr1,Hr2,...,Hrnr],---(10)]]> where Is the receiving spatio-temporal signature of the i-th remote terminal currently active on the channel, as shown in equation (9), nr the total number of remote terminals on the channel.

It should be pointed out that when the channel response length is 1, that is, Mr = 1, and the length of the sliding window is also 1, that is, Lr = 1, the spatio-temporal signature corresponds to our submission on January 20, 1995 entitled "High Spectrum Efficiency and High Efficiency". The spatial signature is described in the pending U.S. Patent Application No. 08/375,848 of "Capacity Wireless Communication System". For this case, the channel response matrix hr is a column vector, and the receiving vector has the following form: zr(t)=hrsr(t-τ)+nr(t). (11)

In a propagation environment with limited delay spread, this is a model suitable for narrowband communication signals. For this situation, we can use the high-capacity wireless communication system described in the pending US Patent Application No. 08/375,848 entitled "Spectrum High-efficiency High-Capacity Wireless Communication System" filed on January 20, 1995. Use spatial processing in the design.

Another special case of the above model is obtained by setting the number of antennas equal to 1, that is, m=1, and the length of the sliding window is also 1, that is, Lr=1. The output of a single receiver is given by the following formula: zr(t)=hr(1)srMr(t-τ)+nr(t).---(12)]]>This model corresponds to the frequently used time spread Discrete time expression of the communication channel. The channel model is established by a finite impulse response filter of length Mr. In this case, time processing can be applied to offset the effect of the communication channel. How to apply this time processing to the case of a single antenna is well known in the art.

The model given in equation (8) combines the spatial and temporal content of the communication channel.

Now consider the communication channel from the base station to the remote terminal. The transmitted spatio-temporal signature is characterized by how the remote terminal receives signals from each antenna array element of the base station on a specific channel. In one embodiment, it is a complex matrix containing the impulse response from the antenna element transmitter output to the remote terminal receiver output as described below.

Let matrix ht be the channel response matrix from the base station transmitter to the remote terminal. The i-th row of the matrix ht is the channel response from the transmitter i to the remote terminal. The maximum length of the channel response is Mt. If the length of the impulse response of the communication channel, antenna element, transmitting and receiving filter is Mt, the impulse response of the channel is equal to the channel response. If the duration of the impulse response is longer, the channel response is an approximation of the impulse response matrix. When the signal st(t) is transmitted from the base station antenna array, let sti(t) be the complex signal transmitted from the i-th base station antenna concurrently st(t)=st1(t)st2(t)...stm(t) .---(13)]]>Now the output received by the remote terminal is given by the following formula: zt(t)zt(t)=[htT(1)···htT(Mt)]stMt(t- τ)+nt(t)---(14)]]>where stMt(t)=st(t)st(tT)...st(t-(Mt-1)T)ht=[ht( 1),ht(2),...,ht(Mt)],---(15)]]>. The term nt(t) represents the noise and interference that appear in the environment and receiver and model errors. Now consider the situation when the signal st(t) at the base station is composed of a scalar signal d(t), st(t)=WtxdLt(t)---(16)]]>Wtx is composed of a complex scalar The Lt×m multiplexing weight matrix, (·)* is the complex conjugate transpose of a matrix, and dLt(t)=d(t)...d(t-(Lt-1)T). ---(17)]]> then maintain the following relationshipUsing the above equation (18) in equation (14), the signal d(t) sent from the base station represents the signal zt(t) received at the terminalDirectly retranslate equation (19) into the following form: zt(t)=wtr*HtdMt+Lt-1(t-τ)+nt(t),---(20)]]>Wherein wtr=&omega ;trT(1)...ωtrT(Lt)is(mLt×1),W~tr=ωtr(1)...ωtr(Lt)is(Lt×m),--- (21)))> andThe transmission space-time signature Ht is an mLt×(Mt+Lt-1) matrix describing the relationship between the space-time transmission vector dMt+Lt-1(t) and the received signal zt(t) at the remote terminal.

When multiple remote terminals are active on the same channel, a single transmission spatio-temporal signature is collected in the multiplexed spatio-temporal signature matrix Ht. For each channel, Ht is formed by using the transmit spatio-temporal signature: Ht=[Ht1,Ht2,...,Htnt],---(23)]]> where Is the transmit spatio-temporal signature of the i-th remote terminal currently active on the channel, as shown in equation (22), nt is the total number of remote terminals on the channel.

It should be noted in the following that assuming the consistency of the sampling time T does not lose its generality.

Figure 2 is a functional block diagram of a multi-channel receiver in a base station.

Figure 3 is a functional block diagram of a space-time demultiplexer in a base station.

Fig. 4 is a functional block diagram of the time filter of the time-space demultiplexer in the base station.

Fig. 5 is a functional block diagram of a space-time multiplexer in a base station in a base station.

Figure 7 is a functional block diagram of a multi-channel transmitter in a base station.

Fig. 8 is a functional block diagram of the spatio-temporal processor in the base station.

Fig. 9 is a functional block diagram of a remote terminal with a repeater switch.

Fig. 10 is a functional block diagram of a remote terminal.

Figure 11 is a schematic diagram of a network system composed of three base stations and a multi-base station controller. Reference Signs Table 1. Base station 2. Base station communication link 3. Base station controller 4. Demodulated received signal 5. Separate uplink signal 6. Received signal measurement 7. Demultiplexed weighting 8. Directional transmission Data 9. Multiplexed modulation signal for transmission 10. Modulation to be transmitted, multiplexed signal 11. Base station calibration signal to be transmitted 12. Multiplexing weighting 13. Space-time processor 14. Multi-channel transmission machine

15. Multi-channel receiver 16a. Multi-channel receiver 16m. Multi-channel receiver 17a. Multi-channel transmitter 17m. Multi-channel transmitter 18a. Transmit antenna 18m. Transmit antenna 19a. Receive antenna 19m. Receive antenna 20. Space-time multiple Path splitter 21. Adder 22a. Time filter 22i. Time filter 22m. Time filter 23. Space-time multiplexer 24. Signal modulator 25. Signal demodulator 26a. Multiplier 26b. Multiplier 26L Multiplier 27a. Sampling delay 27b. Sampling delay 27L. Sampling delay 28. Adder 29a. Time filter 29i. Time filter 29m. Time filter 30a. Multiplier 30b. Multiplier 30L. Multiplier

31a. Sampling delay 31b. Sampling delay 31L. Sampling delay 32. Adder 33. Space-time control data 35. Shared receiver oscillator 36. Receiver control data 37. Transmitter control data 38. Shared transmitter oscillator 39. Space-time Processor controller 40. Active remote terminal table 41. Channel selector 42. Remote terminal database 43. Space-time weighting processor 44. Space-time signature processor 45. Remote terminal antenna 46. Remote terminal duplexer 47. Remote terminal dual Worker output 48. Remote terminal receiver 49. Signal received by the remote terminal 50. Calibration signal received by the remote terminal 51. Remote terminal demodulator 52. Remote terminal demodulated data 53. Remote terminal keyboard and keyboard controller 54. Remote Terminal keyboard data 55. Remote terminal display data 56. Remote terminal display and display controller 57. Remote terminal modulator 58. Remote terminal data to be transmitted

59. Remote terminal modulation data to be transmitted 60. Remote terminal transmitter 61. Remote terminal transmitter output 62. Remote terminal transmitter control data 63. Remote terminal receiver control data 64. Remote terminal microphone 65. Remote terminal microphone signal 66 Remote terminal speaker 67. Remote terminal speaker signal 68. Remote terminal central processing unit 69. Remote terminal repeater switch 70. Remote terminal repeater switch control 71. Wide area network 72. Multi-base station controller 73a. Cellular boundary 73b. Cellular boundary 73c Cellular border 74. High-speed message link 75. Remote terminal input or uplink wireless transmission to an antenna array composed of m receiving antenna elements 19 (a, .., m), each of which is connected to the output To one of m multi-channel receivers in a row of phase-coherent multi-channel receivers 15.

The illustrated embodiment describes a conventional frequency division multiple access (FDMA) system. Each multi-channel receiver can handle multiple frequency channels. The symbol Ncc will be used to mark the maximum number of regular frequency channels that can be processed by the receiver. Depending on the frequency allocated for the operation of the wireless communication system and the bandwidth selected for a particular communication link, Ncc can be as small as 1 (a single frequency channel) or as large as thousands. In an alternative embodiment, the multi-channel receiver 15 can be used instead of processing multi-slots, multi-coding, or some combination of these well-known multiple access techniques. In an alternative embodiment, the channel may consist of multiple regular channels.

In each channel, the receiving antenna elements 19 (a,...,m) each measure the combination of arriving uplink signals from remote terminals sharing the channel. These combinations are generated by the relative position of the antenna elements, the position of the remote terminal, the frequency characteristics of the receiver and transmitter, the spectral content of the signal, and the RF propagation environment and are given by equation (5).

Figure 2 depicts a single multi-channel receiver 16 (a,..., m). The shared local receiver oscillator 35 ensures that the signal from the receiving antenna element 19 (a,..., m) is coherently converted to baseband; the Ncc frequency is set so that the multi-channel receiver 16 (a,..., m) Extract all the Ncc frequency channels of interest. The frequency of the shared local receiver oscillator 35 is controlled by the spatio-temporal processor 13 (FIG. 1) through the receiver control data 36. In an alternative embodiment, multiple frequency channels are all contained in adjacent frequency bands, and a common local oscillator is used to down-convert the entire frequency band that is digitized thereafter. The digital filter and decimator use well-known techniques to extract the desired channel sub-bands. set.

The illustrated embodiment describes an FDMA system. In a TDMA or CDMA system, a shared oscillator 35 should be added to forward a shared time slot or shared coded signal from the spatio-temporal processor 13 to the multi-channel receiver 16 (a, ..., m) through the receiver control data 36, respectively. In these embodiments, the multi-channel receiver 16 (a,..., m) performs the selection of a conventional time division channel or a conventional code division channel in addition to down-conversion to baseband.

Referring again to FIG. 1, the multi-channel receiver 15 generates a received signal measurement value 6 that is provided to the spatio-temporal processor 13 and a set of spatio-temporal demultiplexers 20. In this embodiment, the received signal measurement value 6 includes m complex baseband signals for each of the Ncc frequency channels.

FIG. 8 shows a more detailed block diagram of the spatio-temporal processor 13. The spatiotemporal processor 13 generates and maintains the spatiotemporal signature of each remote terminal for each frequency channel, and calculates spatiotemporal multiplexing and demultiplexing weights for the use of the spatiotemporal demultiplexer 20 and the spatiotemporal multiplexer 23 . In this preferred embodiment, the spatiotemporal processor 13 is implemented using a digital signal processor (DSP) device including a conventional central processing unit. The received signal measurement value 6 enters the spatiotemporal signature processor 44 that estimates and updates the spatiotemporal signature. The spatiotemporal signature is stored in the spatiotemporal signature table in the remote terminal database 42 and used by the channel selector 41 and the spatiotemporal weighting processor 43. The spatiotemporal weighting processor 43 also generates a demultiplexing weight 7 and a multiplexing weight 12. The spatiotemporal processor controller 39 is connected to the spatiotemporal weighting processor 43, and also generates receiver control data 36, transmitter control data 37, and spatiotemporal control data 33.

Referring again to FIG. 8, the spatiotemporal demultiplexer 20 combines the received signal measurement value 6 according to the spatiotemporal demultiplexing weight 7. FIG. 3 shows a space-time demultiplexer 20 for a single channel. In Figure 3, xri represents the i-th element of the reception measurement vector 6 of a single channel, Represents the complex conjugate transpose of the i-th column of the demultiplexing weight matrix of the remote terminal using the channel. W~rx=[ωrx1,ωrx2,...,ωrxm]=ωrx(1)...ωrx(Lr)---(24)]]>Figure 4 illustrates Processing of the i-th received signal xri of a single channel. In this embodiment, a general-purpose arithmetic chip is used to perform arithmetic operations in the time filter 22 (a, ..., m). In Figure 4, xri represents the i-th element of the reception measurement vector 6 of a single channel, Represents the complex conjugate of the j-th element of the i-th vector element of the spatio-temporal demultiplexing weight vector 7 of the remote terminal using the channel. wrx=ωrxT(1)...ωrxT(Lr)ωrx(j)=[ωrx1(j),...,ωrxm(j)]---(25)] ]>For each remote terminal on each channel, the i-th time filter 22i calculates ωrxi*(1)xri(t)+ωrxi*(2)xri(tT)+...+&omega ;rxi*(Lr)xri(tT(Lr-1))---(26)]]>Multiplication is performed by multiplier 26(a,b,...,L), and addition is performed by adder 28 . For each remote terminal on each channel, the spatiotemporal demultiplexer 20 uses the adder 21 to add the output of the time filter 22 (a,...,m), and then outputs wrx*zr(t)-- -(27)]]> For each remote terminal on each channel, the output of the adder 21 given by equation (27) includes the separated uplink signal 5.

Referring again to FIG. 1, the output of the spatiotemporal demultiplexer 20 is a separate uplink signal 5 for each remote terminal communicating with the base station. The separated uplink signal 5 is demodulated by the signal demodulator 25, and a demodulated received signal 4 is generated for each remote terminal communicating with the base station. The demodulated received signal 4 and the corresponding spatiotemporal control data 33 can be provided to the base station controller 3.

In an alternative embodiment, the demultiplexing and demodulation processing are performed together in a non-linear multi-dimensional signal processing unit.

In an embodiment that performs channel coding of the signal sent by the remote terminal, the base station controller 3 sends the demodulated received signal 4 to the spatio-temporal processor 13, and the spatio-temporal processor 13 estimates the bit error rate (BER) using well-known decoding techniques, It is compared with the acceptable threshold value stored in the remote terminal database 42. If the BER is unacceptable, the spatio-temporal processor 13 reallocates resources to alleviate the problem. In one embodiment, except that the current channel is unacceptable, the same strategy as adding a new user is used to allocate links with unacceptable BER to the new channel, unless the current user group of the specific channel changes. In addition, when the channel is available, the remote terminal/base station pair is recalibrated to receive signatures.

For transmission, the signal modulator 24 generates a modulated signal 9 for each remote terminal to which the base station is transmitting, and a set of space-time multiplexing weights 12 for each remote terminal is applied to the modulation in the space-time multiplexer 23 The corresponding delayed version of the signal in order to produce a multiplexed signal 10 to be transmitted for each of the m transmitting antennas 18 (a,...,m) and each of the Ncc channels.

In the illustrated embodiment, the number of downlink channels, Ncc, is the same as the number of uplink channels, Ncc. In another embodiment, the number of uplink and downlink channels may be different. In addition, just as in the case of interactive TV applications where the downlink consists of broadband video channels and the uplink uses narrowband audio/data channels, the channels can be of different types and bandwidths.

In addition, the illustrated embodiment shows the same number of transmit and receive antenna elements m. In other embodiments, the number of transmitting antenna elements and the number of receiving antenna elements may be different, up to and including the case where only one transmitting antenna element in a unidirectional direction is used for transmission, such as in interactive television applications.

Figure 5 shows a space-time multiplexer of a remote terminal on a specific channel. The arithmetic operation in the space-time multiplexer 23 is performed by a general-purpose arithmetic chip. Let d(t) denote the element of the modulation signal 9 defined for the remote terminal on the channel, and denote the multiplexing weight vector by wtx. The multiplexing weight vector wtx is related to the multiplexing weight matrix Wtx by the following formula. wtx=ωtxT(1)...ωtxT(Tt)is(mLt×l)W~tx=ωtx(1)...ωtx(Lt)is(Lt×m),- --(28)]]>

For each remote terminal on each channel, the space-time multiplexer 23 calculates the product of its multiplexing weighting matrix 12 and the delay pattern of the modulated signal d(t)9: st(t)=WtxdLt( t)=ωtx(1)...ωtx(Lt)*d(t)...d(t-(Lt-1)T)---(29)))> Time Filter 29 (a,...,m) Calculate the product of the row of the multiplexing weighting matrix and the delay pattern of the modulated signal d(t)9. For each remote terminal on each channel, the i-th time filter 29i calculates ωtxi*(1)d(t)+ωtxi*(2)d(tT)+...+ωtxi *(Lt)d(tT(Lt-1))--(30)]]>The multiplication operation is performed by the multiplier 30 (a, b,..., L), and the addition operation is performed by the adder 32. For each channel, equation (29) is estimated by the space-time multiplexer 23 of each remote terminal transmitting on that channel. Corresponding to each remote terminal are different multiplexing weighting vectors, multiplexing weighting matrices, and modulation signals. For each channel, the spatio-temporal multiplexer 23 adds up the multiplexed signals of each remote terminal being transmitted to it on that channel, and generates as the signal to be transmitted from each antenna for each downlink channel. Signal modulation and multiplexing signal 10, st(t).

The modulated and multiplexed signal 10 is input to a row of m-phase coherent multi-channel transmitters 14. Figure 7 depicts a multi-channel transmitter 17 (a,...,m) with an antenna connection, a shared local transmitter oscillator 38, and a digital input 10. The shared local transmitter oscillator 38 ensures that the relative phase of the multiplexed signal 10 is maintained by the transmitting antenna 18 (a, ..., m) during transmission. The space-time processor 13 (see FIG. 1) controls the frequency of the shared local transmitter oscillator 38 through the transmitter control data 37.

In an alternative embodiment, the spatio-temporal multiplexer 23 uses the well-known baseband multiplexing technique to multiplex all the calculated channel signals to be transmitted into each multi-channel transmitter 17 (a,. .., m) Up-converted and transmitted single bandwidth signal. Multiplexing is performed in digital or analog mode as required.

The illustrated embodiment shows a system with multiple frequency channels. In a time division multiple access or code division multiple access system, a shared oscillator 38 is added to transmit the shared time slot from the space-time processor 13 to the multi-channel transmitter 17 (a,..., m) through the transmitter control data 37. Or share the coded signal.

Referring again to FIG. 1, in an application that needs to transmit a spatio-temporal signature, the spatio-temporal processor 13 can also transmit a predetermined calibration signal for each antenna on a specific downlink channel. The spatio-temporal processor 13 instructs the multi-channel transmitter 17 (a,...,) to transmit a predetermined calibration signal 11 through the transmitter control data 37 to replace the multiplexed signal 10 of a specific downlink channel. This is a mechanism used to determine the transmission spatio-temporal signature of the remote terminal on the downlink channel.

In an alternative embodiment that uses well-known channel coding techniques to encode the signal to be transmitted to the remote terminal, the remote terminal uses well-known decoding techniques to estimate the BER that is thereafter reported back to the base station on its uplink channel. If these BERs exceed acceptable limits, take corrective action. In one embodiment, in addition to being unable to accept the current channel, the correction activity involves reallocating resources using the same strategy as adding new users, unless the current user group of that particular channel changes. In addition, the transmission signature of the remote terminal/base station pair is recalibrated when the channel is available.

Figure 9 depicts the arrangement of components in a remote terminal that provides voice communication. The antenna 45 of the remote terminal is connected to the duplexer 46 to allow the antenna 45 to be used for transmission and reception. In an alternative embodiment, separate receive and transmit antennas are used without the need for a duplexer 46. In another alternative embodiment, to receive and transmit on the same frequency channel but at different times, it is well known to use a transmit/receive (TR) switch instead of a duplexer. The output 47 of the duplexer serves as the input to the receiver 48. The receiver 48 generates a down-converted signal 49 and inputs it to the demodulator 51. The demodulated received voice signal 67 is input to the speaker 66.

The demodulated reception control data 52 is provided to the central processing unit 68 (CPU) of the remote terminal. The demodulated reception control data 52 is used to receive data from the base station 1 during call establishment and termination, and in an alternative embodiment, is used to determine the quality (BER) of the signal received by the remote terminal and send it back as described above Base station.

The remote terminal CPU68 is realized by a standard DSP device. The remote terminal CPU 68 also generates receiver control data 63 for selecting the receiving channel of the remote terminal, transmitter control data 62 for setting the transmission channel and power level of the remote terminal, control data 58 to be transmitted, and The display data 55 of the remote terminal display 56. The remote terminal CPU 68 also receives keyboard data 54 from the remote terminal keyboard 53.

The remote terminal voice signal 65 to be transmitted from the microphone 64 is input to the modulator 57. The remote terminal CPU 68 provides control data 58 to be transmitted. The control data 58 to be transmitted is used to transmit data to the base station 1 during call establishment and termination, and to transmit information during calls such as measuring call quality (e.g., bit error rate (BER)). The transmitter 60 up-converts and amplifies the modulated data 59 to be transmitted output by the modulator 57 to generate a transmission output signal 61. The transmitter output 61 to be transmitted is then input to the duplexer 46 through the antenna 45.

In an alternative embodiment, the remote terminal provides digital data communication. The demodulated received voice signal 67, the speaker 66, the microphone 64, and the voice signal 65 to be transmitted are replaced with a digital interface known in the art that allows data to be transmitted to and from an external data processing device (such as a computer).

Referring again to FIG. 9, the remote terminal allows the receiving data 49 to be transmitted back to the base station 1 through the switch control signal 70 via the switch 69 controlled by the remote terminal CPU 68. In normal operation, the switch 69 drives the transmitter 60 with the modulation signal 59 of the modulator 57. When the base station 1 commands the remote terminal to enter the calibration mode, the remote terminal CPU 68 sends a predetermined calibration signal 58 to the modulator 57 instead of the remote terminal microphone signal 65. This is the mechanism used to determine the received spatio-temporal signature of the remote terminal on the uplink channel. In the repeater mode, the remote terminal CPU 68 triggers the switch control signal 70, and the remote terminal CPU 68 commands the switch 69 to drive the transmitter 60 with the received data 49.

Figure 10 shows an alternative embodiment of the remote terminal repeater function. The switch 69 of Fig. 9 is no longer used. Instead, the output of the receiver 48 is provided to the remote terminal CPU 68 via the data connection 50. In normal operation, the data connection 50 is ignored in the remote terminal CPU 68. In the calibration mode, the remote terminal CPU 68 uses the data link 50 to calculate the remote terminal's transmission spatio-temporal signature, and then transmits it back to the base station 1 through the modulator 57 and the transmitter 60 as the control data 58 to be transmitted.

In an alternative embodiment, a spatio-temporal calibration procedure in the remote terminal is not required. In many conventional wireless protocol standards, remote terminals regularly report received signal strength or received signal quality to the base station. In this embodiment, the received signal strength report is sufficient to calculate the transmission spatiotemporal signature of the remote terminal, as described below. General Principle-Base Station In many aspects, the performance of the spectrum-efficient base station shown in Figure 1 is very similar to that of a standard wireless communication system. The main difference is that: Compared with conventional base stations that use the same time/frequency resources, a spectrum-efficient base station supports more simultaneous calls. The communication channel can be a frequency channel, a time channel, a coding channel, or any combination of these channels. The spatio-temporal multiplexer/demultiplexer increases system capacity by allowing multiple simultaneous communication links on each of these channels. In addition, by combining signals from multiple receiving antennas, the spatio-temporal demultiplexer 20 generates a separate uplink signal 5 that is equalized simultaneously in space and time. The separated uplink signal 5 will be the result of having substantially improved signal-to-noise ratio, reduced interference, and improved quality in a multipath environment compared to standard base stations.

In the illustrated embodiment, a wireless communication system composed of a plurality of remote terminals and a base station combining antenna array and spatiotemporal signal processing is described. The system has applications such as providing wireless access to the local PSTN. The initial information transfer (or call) is initiated by a remote terminal or by the communication link 2 through the base station controller 3. As is well known in the art, call initialization is performed on the downlink and uplink control channels. In this embodiment, the transmission antenna 18 (a, ..., m) is used to transmit the downlink control channel. In an alternative embodiment, the downlink control channel is broadcast from a single omnidirectional antenna. The base station controller 3 transmits the identification of the remote terminal involved in the call to the spatio-temporal processor 13 using the stored spatio-temporal signature of the remote terminal to determine which communication channel the remote terminal should use. The selected channel may have been occupied by several remote terminals, however, the spatiotemporal processor 13 uses the spatiotemporal signature of all remote terminals on the channel to determine that they can share the channel without interference.

The spatiotemporal processor 13 uses the spatiotemporal multiplexing and demultiplexing weights calculated for the selected channel and the above-mentioned remote terminal to form the spatiotemporal multiplexer 23 and the spatiotemporal demultiplexer 20. Then, the spatiotemporal processor 13 notifies the controller 3 of the selected channel. As in a conventional base station, the controller 3 then commands the remote terminal (via the downlink control channel) to switch to the selected channel to continue communication. In the case that the remote terminal has power control capabilities, as is well known in the art, the controller 3 also commands the remote terminal to rely on, for example, the power level of other remote terminals sharing the same channel and the signals required for each link as described below. Parameters such as quality adjust its power to an appropriate level. When the communication ends, the remote terminal returns to its idle state in which it monitors the downlink control channel and waits for its next call. Spatio-temporal processing-base station FIG. 8 shows a block diagram of the spatio-temporal processor 13. The spatiotemporal processor controller 39 controls the spatiotemporal processor 13, and the spatiotemporal processor controller 39 is connected to the base station controller 3 via a link 33. The spatio-temporal processor controller 39 controls the gain and frequency settings of the multi-channel transmitter 14 and the multi-channel receiver 15 through the control lines 37 and 36.

The spatio-temporal processor 13 maintains an active remote terminal table 40 listing which remote terminal is currently using each communication channel and its current transmit power level. Other parameters of the remote terminal such as the currently used modulation format, the noise level of the receiver in the current frequency channel, and the current signal quality requirements are also stored. The spatio-temporal processor 13 also maintains a spatio-temporal signature table in the remote terminal database 42. In an alternative embodiment, the remote terminal database 42 includes the power control level of the remote terminal, the allowed transmission and reception frequency channels, and familiar Modulation format table.

The spatiotemporal signature table in the remote terminal database 42 contains the transmission spatiotemporal signature Ht and the reception spatiotemporal signature Hr of each channel used for each remote terminal's operation. In another embodiment, for example, by storing the largest singular value and the corresponding singular vector of the signature, a set of basic vectors spanning the same column space as the transmitting and receiving spatio-temporal signature is stored. In another embodiment, the parameters from which signatures can be formed are stored, for example, by storing transmission and reception channel response matrices. In another embodiment, an estimated value of the quality of the spatiotemporal feature marker (for example, estimated error covariance) is also stored. In yet another embodiment, a parameter describing the uncertainty caused by the temporal change of the spatiotemporal signature is also stored. The transmission spatio-temporal signature includes the influence of the propagation environment between the base station and the remote terminal, and any difference in the frequency characteristics of the transmitter 14, antenna cable, and transmitting antenna 18 (a, ..., m).

When the base station controller 1 forwards a call initialization request of a specific remote terminal via the link 33, the channel selector 41 searches the active remote terminal table 40 to find a communication channel that can accommodate the remote terminal. In the preferred embodiment, there is a receiving active remote terminal table and a transmitting active remote terminal table used by the channel selector 41 to form a multiplexed space-time matrix and a multiplexed space-time matrix of each channel. For each channel, a multiple is formed by adding the received and transmitted spatiotemporal signature of each remote terminal that is currently active on (using) the channel plus an additional part containing the appropriate spatiotemporal signature of the remote terminal requesting the communication channel. The path decomposition and multiplexing spatio-temporal signature matrix, see equation (10-23).

The channel selector 41 calculates the function of these characteristic marker matrices to evaluate whether the communication between the base station and the new remote terminal can be successfully performed on the selected channel. In a preferred embodiment, the channel selector 41 first calculates the space-time multiplexing and demultiplexing weights of the remote terminal, and then uses these weights to estimate the link performance.

In the illustrated embodiment, the space-time multiplexing weight vector is the column of the matrix Wtx, and the k-th column of Wtx is given in equation (31): {Wtx}k=skt{Ht(Ht*Ht)- 1}mLt(k-1)+1---k=1,...,nt,---(31)]]> where (·)-1 is the inverse matrix of the matrix and {·}k is the matrix In the k-th column of, Ht is the multiplexed spatio-temporal signature matrix related to the relevant channel, Is the amplitude of the k-th signal to be transmitted. If the resulting multiplexing matrix is unstable, a stable approximation of the multiplexing matrix is formed using techniques well known in the art. In the preferred embodiment, the mean square noise voltage (Nk) of the remote terminal receiver and the lowest expected signal quality are used Calculate the amplitude to be emitted As given in equation (32): skt=(SNRkdes×Nk)1/2.---(32)]]> Now the channel selector 41 calculates the average mean square voltage to be transmitted from each element (power ) Pt, as the weighted sum of squares of appropriate elements Pt~=diag([W~tx1*···W~txn1*]W~tx1...W~txn1),---(33)] ]>where Is the weighting matrix of the k-th user as defined in equation (21), and diag(·) is the vector obtained by the diagonal elements of the stacked matrix. Calculate the peak squared voltage (power) to be emitted from each element As the square of the appropriately weighted amplitude sum Ptpeak=diag(abs(W~tx1*...W~txnt*)abs(W~tx1...W~txnt)),---(34)))> Where abs(·) is the absolute value of the element (element-wise). The channel selector 41 compares these values with the limits of each element of each transmitter. If either the average or the peak value exceeds the acceptable limit, the above-mentioned remote terminal is not allocated to the candidate channel. Otherwise, the ability to successfully receive from the remote terminal is checked.

In an alternative embodiment, the transmitter limit is used as an inequality constraint in an optimization algorithm that calculates the transmit weight that meets the given technical requirements and results in the smallest amount of possible transmit power. If the transmission weight that satisfies these constraints cannot be found, the above-mentioned remote terminal is not allocated to the candidate channel. This optimization algorithm is well known.

In an alternative embodiment using Time Division Duplex (TDD), due to the assumption that the channel and the interference are reciprocal, the multiplexing weighting is selected as the demultiplexing weighting scaling type. The calibration parameters are selected to provide sufficient SINR at the remote terminal.

To test the uplink, the channel selector 41 uses the demultiplexed spatiotemporal signature matrix Hr related to the relevant channel to calculate the spatiotemporal demultiplexing weight Wrx. In the illustrated embodiment, the spatiotemporal demultiplexing weight vector is the column of the matrix Wrx given in equation (35): Wrx=(HrPrHr*+Rnn)-1HrP-r,---(35)]] > Where the k-th column of Pr is given by equation (36) {Pr}k=(Pr)Lr(k-1)+1---k=1,...,nr,---(36) ]]>where Pr is the signal transmitted by the remote terminal The (diagonal) matrix of the mean square amplitude (power) Pr=E(sr1,Lr(t)...srnr,Lr[sr1Lr*(t)···srnr,aLr*]) --- (37)]]>Rnn=E{er(t)er*(t)} is the base station noise covariance. Then, in one embodiment, the expected value of the normalized mean square error covariance is calculated as follows: MSE~=Pr-1/2*((I-Wrx*Hr)Pr(I-Wrx*Hr)*+Wrx* RnnWrx)Pr-*/2---(38)]]> where the symbol (·)-*/2 represents the complex conjugate transpose of the square root of the matrix. The inversion of MSE is the estimated value of the expected signal-to-interference plus noise ratio (SINR) output by the space-time demultiplexer: SINR = MSE-1(39)

If all diagonal elements of the SINR are above the desired threshold based on the required signal quality received from each remote terminal, then the remote terminal is allowed to access the channel. If the candidate remote terminal is below its threshold and has the ability to increase its output power, the same calculation is performed again to increase the power output of the remote terminal until the maximum output power of the remote terminal is reached and the SINR is still not suitable. If possible, In the case of increasing its power, the SINR of another remote terminal is below its threshold or exceeds all thresholds. If an acceptable remote terminal transmit power can be found, the remote terminal is allowed to access the specific channel, otherwise the access is denied and another channel is checked.

In an alternative embodiment, a well-known optimization procedure is used to calculate the demultiplexing weight, with the goal of minimizing the transmission power of the remote terminal so that the estimated signal at the base station meets or exceeds its minimum expected SINR.

In addition, in an alternative embodiment, in the event that a channel that can accommodate a remote terminal is not found, the channel selector 41 calculates whether certain rearrangements of existing remote terminals in the channel allow remote terminals to be supported on certain channels. In this case, if the existing user's rearrangement does not allow the remote terminal to be accommodated, the remote terminal communication is only rejected at that time.

In an alternative embodiment employing frequency division duplex (FDD), there is no restriction on the allocation of fixed channel pairs for transmission and reception to remote terminals. With a sufficiently flexible system configuration, the channel selector 41 can select a specific remote terminal to allocate to transmit and receive channels separated by different frequency duplex offsets, so as to minimize the interference level of the entire system.

Since adding new remote terminals to the channel can change them significantly, the space-time multiplexing and demultiplexing weights of remote terminals that have already used a channel must be recalculated. In the preferred embodiment, the channel selector 41 that has performed the required calculations sends the new spatiotemporal multiplexing and demultiplexing weights to the spatiotemporal weighting processor 43 for establishing the spatiotemporal multiplexer 23 and demultiplexer 20. In an alternative embodiment, the spatiotemporal weighting processor 43 uses the spatiotemporal signature matrix sent to it by the channel selector 41 to calculate different spatiotemporal multiplexing and demultiplexing weighted sets for all remote terminals on the channel. .

Then, the spatiotemporal weighting processor 43 sends the new spatiotemporal demultiplexing weight of the channel to the spatiotemporal demultiplexer 20, and sends the new spatiotemporal multiplexing weight of the channel to the spatiotemporal multiplexer 23 to update the active remote The terminal table 40 notifies the spatiotemporal processor controller 39, and the spatiotemporal processor controller 39 notifies the base station controller 3 of the selected channel. Then, the base station controller 3 transmits a message to the remote terminal using the downlink control channel that instructs the remote terminal to switch to the desired channel.

From equation (31), it can be shown that the multiplexing weight Wtx has the characteristic: Wtx*Ht=diag(st1,···,stnt), ---(40)]]> where diag(·) Is a diagonal matrix with diagonal elements formed by vectors. This means that at the kth remote terminal, the signal to be sent to the terminal is received with a sufficient (positive real number) amplitude stk. The case with zero off-diagonal elements means that at the kth remote terminal, the remote terminal does not receive other transmitted signals. In this way, each remote terminal only receives signals for its use at the required power level to ensure correct communication. In an alternative embodiment, the uncertainty in the estimated value of Ht is added to setting the base station transmit power level and calculating the weight to minimize the influence of the error and/or the change in Ht.

Similarly, in this base station, the specific demultiplexing weights given in equation (35) have characteristics conditioned on the received spatio-temporal signature and the data of the voltage (power) transmitted from the remote terminal, which is given by the following equation Estimate the signal S^=Wrx*zr,---(41)]]> is the most accurate in the sense of minimum mean square error. In particular, they most closely match the signal transmitted by the remote terminal, giving measurements calculated by multiple antenna elements at the base station.

Equations (31) and (35) only represent one way to calculate the weights of space-time multiplexing and demultiplexing. There are other similar strategies to prove similar characteristics to those shown in equation (40) and those described in the previous paragraph. Other well-known techniques for calculating weighted Wtx and Wrx account for the uncertainty in multiplexing and demultiplexing the spatiotemporal signature matrix Ht and Hr, and can introduce more complex power and dynamic range constraints. Determining the spatio-temporal feature mark is shown in FIG. 8. The spatio-temporal processor 13 also includes a spatio-temporal feature mark processor 44 for finding the spatio-temporal feature mark of the remote terminal. In the illustrated embodiment, the spatiotemporal signature processor 44 uses a calibration technique similar to that described in our US Patent 5,546,090 (published on August 13, 1996) entitled "Method and Apparatus for Calibrating Antenna Array".

In the illustrated embodiment, each remote terminal can enter the calibration mode. In the calibration mode, the remote terminal can transmit a predetermined signal, and can also enter a repeater mode that sends the received signal 49 back to the base station 1. The calibration mode is controlled by the remote terminal CPU68. Referring to FIG. 9, the repeater mode is provided by the switch 69 controlled by the remote terminal CPU 68 through the switch control signal 70.

In order to determine the transmission and reception spatio-temporal signature of the remote terminal, the spatio-temporal signature processor 44 commands the remote terminal to enter the calibration mode by transmitting a command to it on the downlink channel. The command is generated by the base station controller 3 according to a request from the spatio-temporal processor controller 39 and modulated by the signal modulator 24. In an alternative embodiment, the calibration mode is entered regularly in predetermined stages.

Then, the remote terminal transmits a predetermined terminal calibration signal on the channel. In this embodiment, the terminal calibration signal is a known pseudo-random noise sequence limited within the current channel frequency band. In another embodiment, the predetermined terminal calibration signal is any known signal. In accordance with equation (5) and in the absence of noise and parameter offset, the time sample X of the received data stored in the m×Nr data matrix X given by the following equation X = hrSr (42) where Sr is the predetermined terminal calibration The Mr×Nr matrix of the signal. The receive channel response matrix is given by

Among them, Sr is the well-known Moore-Penrose pseudo-inverse matrix of matrix Sr. For a full-level matrix Sr with more columns than rows, it satisfies SrSr = I (the same matrix), and for a full-level matrix Sr with more rows than columns, it satisfies Sr Sr=I. The receiving spatio-temporal signature Hr can now be constructed from the channel matrix hr.

In an alternative embodiment, the reception is directly determined by storing the time samples of the space-time reception vector in an mLr×Nr data matrix Z according to equation (8) and without noise and parameter offset. The spatio-temporal signature of HrZ=HrSr (44) where, in this case, Sr is the Mr+Lr-1×Nr matrix of the predetermined terminal calibration signal. The received spatio-temporal signature is given by the following formulaWhen other terminals are using the same channel, the receiving spatio-temporal signature can be determined.

In an alternative embodiment, the spatiotemporal demultiplexing weight vector is directly determined from equation (46): wrx*Z=Sr.---(46)]]> where in this case, Sr contains the predetermined terminal calibration signal 1×Nr vector. The weighted vector of space-time multiplexing is given byWhen other terminals are using the same channel, the receiving spatio-temporal demultiplexing weight vector can be determined.

In an alternative embodiment, the terminal calibration signal is not all known, but has parameters of a predetermined modulation format. Several techniques known in the art use parameters of a predetermined modulation format, such as a constant modulus signal, to determine the receive channel matrix. Once the receiving channel matrix is determined, the receiving spatio-temporal signature is formed as described above. It is also clear to those skilled in the art how to use the predetermined modulation format parameters of the terminal calibration signal to determine the received spatiotemporal signature or demultiplexing weight vector.

In alternative embodiments related to those described in U.S. Patent No. 5,546,090 (published on August 13, 1996) entitled "Method and Apparatus for Calibrating Antenna Arrays", well-known techniques are used to explain the noise and vibrations that occur in the system. Parameter changes such as the frequency offset of the converter.

Once Hr is known, the demultiplexing weight is calculated, and the remote terminal enters the repeater mode. Then, the spatio-temporal signature processor 44 transmits a predetermined base station calibration signal on the channel occupied by the remote terminal by instructing the multi-channel transmitter 17 (a, ..., m) and the spatio-temporal processing controller 39 by means of the transmitter control data 37 11. In this embodiment, the m signals (corresponding to each antenna) in the predetermined base station calibration signal 11 are different from the known pseudo-random noise sequence defined in the current channel frequency band. In another embodiment, the predetermined base station calibration signal 11 is any known, different signal.

The remote terminal shown in FIG. 9 sends back the signal received at the remote terminal. The multi-channel receiver 15 in the base station 1 shown in FIG. 1 receives the forwarded signal and provides it to the spatio-temporal signature processor 44 shown in FIG. 8. The time sampling of the received data is processed by demultiplexing and weighting, and the resulting signal is stored in the 1×Nt data matrix Zt given by the following equation according to equation (14) and without noise and parameter offset. Zt= k[htT(1)···htT(Lt)]St---(48)]]> where St is the mLt×Nt matrix of the predetermined base station calibration signal, k is the known quantity, and the signal is sending back Amplify the known quantity in the remote terminal before reaching the base station.

The transmit channel response is given byThe transmission spatio-temporal signature Ht can now be formed from the transmission channel response.

In an alternative embodiment, the base station calibration signal is not fully understood, but has the parameters of the predetermined modulation format. Several known techniques in the art use predetermined modulation format parameters to determine the transmission channel matrix.

Also in the alternative embodiment described in U.S. Patent No. 5,546,090 (published on August 13, 1996) entitled "Method and Apparatus for Calibrating Antenna Arrays", well-known techniques are used to explain noise and oscillator frequency deviations that occur in the system. Parameter changes such as shift.

In an alternative embodiment, the calibration mode includes only the above-mentioned transponder mode. The receiving channel matrix is then determined by one of the several techniques described in the following documents. For examples, see IEEE Transaction on Signal Processing by E. Moulines, P. Duhamel, J.-F. Cardoso, and S. Mayrargue in February 1995. 43(2): "Subspace Method for Blind Identification of Multi-channel FIR Filters" published by 516-525.

The spatiotemporal signature processor 44 stores the new spatiotemporal signature in the remote terminal database 42. Upon completion, the spatio-temporal signature processor 44 orders the remote terminal to exit the calibration mode by transmitting a command on the downlink channel to the remote terminal.

In an alternative embodiment, the remote terminal can directly perform the calculation of the spatiotemporal signature of the remote terminal's transmission. This embodiment of the remote terminal is shown in FIG. 10. In the calibration mode, as before, the spatio-temporal signature processor 44 transmits a predetermined calibration signal 11 on the channel to be calibrated by the remote terminal. The remote terminal CPU 68 uses the same technology used by the spatiotemporal signature processor 44 in the previous embodiment to use the received calibration signal 50 and the known transmission waveform to calculate the remote terminal's transmission spatiotemporal signature. The calculated transmission spatio-temporal signature is sent back to the base station 1 through the modulator 57 and the transmitter 60 as the control data 58 to be transmitted. When the base station 1 receives the spatiotemporal signature, the spatiotemporal signature processor 44 stores the new transmission spatiotemporal signature in the remote terminal database 42. Since each remote terminal performs the calculation of the transmission spatio-temporal signature separately, this arrangement allows multiple remote terminals to simultaneously calculate their own transmission spatio-temporal signature on the same channel. In this embodiment, the spatiotemporal signature processor 44 calculates the spatiotemporal signature received by the remote terminal in the same manner as the previous embodiment.

Using these technologies, the spatiotemporal signature processor 44 can transmit and receive spatiotemporal signatures for a specific channel at any time when the channel is idle. The efficiency of these calibration techniques allows the spatiotemporal signature processor 44 to update the spatiotemporal signature of a large number of remote terminals for a particular channel when the channel is occupied for only a short time.

In an alternative embodiment, the received spatio-temporal signature is obtained in a decision feedback mode. The received data is demodulated and re-modulated to produce an estimate of the original modulated signal. These techniques allow estimation of the received spatio-temporal signature even when multiple remote terminals are occupying a single channel.

In another embodiment, the system can be designed to continuously update the spatiotemporal signature of the remote terminal in a "closed loop" manner. This is done to account for the temporal changes of the spatiotemporal signature due to, for example, the movement of the remote terminal or the change in RF propagation conditions. To this end, both the base station and the remote terminal periodically transmit predetermined training sequences. A different predetermined training sequence is allocated to each remote terminal currently active on the specific channel, and the training sequence is given to all other remote terminals currently active on the specific channel. In one embodiment, in the sense that the inner product of any two training sequence waveforms is zero, the different training sequences are orthogonal. Whenever a training sequence is transmitted, each remote terminal uses well-known techniques to calculate how much of each training sequence it has received, and transmits this information to the base station.

In the illustrated embodiment, the base station uses the receiver output and the learned transmission waveform to calculate the remote terminal reception spatio-temporal signature. In another embodiment, the base station calculates how much of the training sequence transmitted by each remote end has been experienced on each output of the space-time demultiplexer, expressed as a complex coupling coefficient. Knowing these coupling coefficients allows correction of the currently active receive and transmit spatio-temporal signatures so that well-known techniques can be used to reduce mutual interference.

Finally, in a full-duplex communication system using Time Division Duplex (TDD), as is well known in the art, the transmitting and receiving frequencies are the same. In this case, the well-known principle of reciprocity is used to directly correlate the transmitted and received space-time signatures. Therefore, this embodiment only determines one of the signatures, for example, the received spatiotemporal signature, the other signature, in this case, the spatiotemporal signature is transmitted, from the first (received) spatiotemporal signature and the known relative phase and The amplitude characteristics of the multi-channel receiver 15 and the multi-channel transmitter 14 are calculated. Network level spatio-temporal processing In the embodiment described here, the spatio-temporal processor of each base station in the cellular wireless communication system works independently to maximize the number of communication channels in the direct cellular. However, if the spatiotemporal processors from each base station communicate with spatiotemporal processors from other neighboring cells and coordinate their effects with the spatiotemporal processors, a significant improvement in system capacity can be achieved. Figure 11 shows a specific embodiment.

The multi-base station controller 72 serves as an interface between the wide area network 71 via the link 74 and the base station 1 (a, b, c) via the base station communication link 2 (a, b, c). Each base station is responsible for providing coverage for many remote terminals. In one embodiment. Each remote terminal is assigned to only one base station, thus defining the cell boundary 73 (a, b, c) in which all remote terminals attached to a particular base station are located. The user equipped with the remote terminal 75 is identified by the box "R" in the figure.

Each spatio-temporal processor included in the base station 1 (a, b, c) measures and stores the spatiotemporal signatures of the remote terminals in its cell and the remote terminals in adjacent cells. The multi-base station controller 72 coordinates the determination of the spatio-temporal signature of remote terminals in adjacent cells through the base station communication link 2 (a, b, c). Through the base station communication link 2 (a, b, c) and the multi-base station controller 72, the spatiotemporal processors in the base stations 1 (a, b, c) from neighboring cells inform each other which remote terminal is communicating on which channel. Each spatio-temporal processor includes spatio-temporal signatures of remote terminals currently active in neighboring cells to form extended multiplexing and demultiplexing spatio-temporal signature matrices Ht and Hr sent to all neighboring base stations. The channel selector in each base station uses these extended spatiotemporal signature matrixes to jointly allocate remote terminals to each channel in each base station 1 (a, b, c).

Use the extended multiplexing and demultiplexing feature mark matrices Ht and Hr to calculate the multiplexing and demultiplexing weights Wtx and Wrx obtained for each base station. In calculating the weight, its purpose is to minimize the signals transmitted to and received from the active remote terminals of the neighboring cells, thereby allowing more remote terminals to communicate at the same time.

In an alternative embodiment, the multi-base station controller 72 uses the active remote terminal/base station/channel link table, the related remote terminal database, and the specific request for the link to be allocated to dynamically allocate remote terminals requesting access to the base station. . In addition, the remote terminal may use multiple (directional) transmitting and receiving antennas to directionally link to multiple nearby base stations according to the instructions of the multiple base station controller 72, thereby further increasing the system capacity. Advantages The apparatus and method according to the present invention provide a significant advantage compared with the prior art in that it allows many remote terminals to share the same communication channel at the same time through simultaneous spatio-temporal multiplexing/demultiplexing. In addition, compared with standard base stations, signals received from and transmitted to remote terminals have a sufficiently improved signal-to-noise ratio, reduced interference, and improved quality in a multipath environment.

Therefore, the wireless communication system can support more calls with the same amount of spectrum, or have a larger data throughput. As an alternative, the wireless communication system can support the same volume of calls or data throughput with a small amount of spectrum. Alternative embodiment In an alternative embodiment, a single array of m antennas replaces the transmitting antenna 18 (a,..., m) and the receiving antenna 19 (a,..., m) of the base station 1 with a single array of m antennas. Each element in the array is connected to its corresponding part of the multi-channel transmitter 14 and its corresponding part of the multi-channel receiver 15 by means of a duplexer.

In another alternative embodiment, the spatio-temporal processing described in pending patent application 07/806,695 can be used to process the signals in the uplink control channel in real time. This allows multiple remote terminals to request a communication channel at the same time.

In another embodiment of data transfer applications involving bursts or data packets, a separate uplink control channel is not required, and the system can service communication requests and other control functions during control time intervals dotted by communication intervals.

Although the above description contains many features, these features should not be construed as limiting the scope of the present invention, but only as an example of a preferred embodiment of the present invention. Many other improvements can be made. Therefore, the scope of the present invention should not be determined by the illustrated embodiments, but should be defined by the appended claims and their legal equivalents.