Method and arrangement for multiplexing several users to the communication channels of a tdma system

The invention relates to a method and associated apparatus for multiplexing telecommunication connections in telecommunication systems based on nap TDMA. A burst structure is defined, the best symbol and the filling a time slot at a radio interface. The burst structure is filled with symbols, forming a transfer burst that is sent within a time slot. Stutters on the threads of the structure of the file with symbols comprise subsets of the threads of the information system symbol (302, 802, DATA) of the cut, and of the information symbol of a second type is taken and with these a second part is filled ( 306, DATA) of the cut. In addition, Control Symbols are taken Flow Control To Fill These Control Parts (301, 303, 304, 305, 307, 601,602, 603, 801, 803, 804, 805, 807, 811,813, 814, 815, 817, 1011, 1012 , 1013, S) of the cut.

datalAhoei jPXfil7EETTIK001M /> 4 7 j VUOWtAW) | li²<tl

for channel coding

DATALAHDE2 ~~ 1 z ραμτεειτικοοοαηΛ 'YT-^B

for channel coding

220

221

222

223 ! 4 u

li

L ·²²'

interleaver

<td>j Burst Shaping Area |</td>

<td>Φ</td>

<td>J MOOULMTTORI |</td>

<td>T - -</td>

<td>J TRANSMITTER I</td>

<td> 1</td>

<td>4 ANTENNI F</td>

<td></td>

<td>ri ANTENNI F</td>

<td> *</td>

<td>J ANTENNA |</td>

<td> 4 ¹</td>

<td>J DEMODULATOR |</td>

<td>Φ</td>

<td>] PURSKEENPURXAIN |</td>

<td> 4 4</td>

ISO DISCOVERY1 —τ “U ²⁵U

LOUIJWSENPURKAIn] - 26 | KWAVADEKQODgffl [jKAMVAPEÄOOPEftj] -&gt;

VA $ TAMUOKKA1N \

Reply / OWCAW

PURPOSE OF THE PARITY i ΡΑΡίΤΕΕΠΝΓΑΡΚ I¹ 'ϊ' ^k

260

261

U⁶² ss

Method and System for Interleaving Multiple Users on TDMA System Communication Channels - Method and Procedure for Attaching Multiplexer Flera Användare Account Communications Communications Channel to TDMA System

The invention relates generally to a technique for allocating time and frequency resources of a radio motion system among system users. In particular, the invention relates to a technique for efficiently utilizing time slots defined in a time division multiple access system in a situation where users may have different and / or dynamically changing needs with respect to the data rate of the radio interface.

In TDMA-based radiotelephone systems, the base station allocates periodically occurring time slots for use by handheld terminals from certain transmission frames. Examples include the well-known GSM (Global System for Mobile Telecommunication) system and its upgraded version EDGE (Enhanced

Datarates for GSM Evolution).

The background of the invention will be briefly described with reference to the known GSM transmission chain with reference to Fig. 1. As an example of a typical circuit-switched service, full-speed voice communication is used herein. The speech recorded by the microphone 101 is first encoded in a speech encoder 102 which converts the analog pu20 signal into a digital format and performs a plurality of encoding operations. The output rate of the speech encoder is 13 kbit / s and consists of 260 bit blocks which follow each other at 20 milliseconds. Channel encoder 103 adds redundancy to this data stream, increasing its rate by adding information calculated from the contents of the blocks. The purpose of channel coding is to enable detection and even correction of signal errors occurring later during transmission. The output of the channel encoder 103 consists of codewords of 456 bits each. Each input data block entering the channel encoder will consist of exactly one code word.

The codewords from the channel encoder 103 are fed to the interleaver / burst shaping unit 104 for mixing bits of a plurality of codewords in a predetermined manner and arranging them into bursts. The purpose of interleaving is to de-correlate the error rates of any errors in the transmission so that the resulting erroneous bits are distributed at random points in a plurality of code words instead of corrupting the sequence of consecutive bits in a single code word.

Most of the currently used interleaving methods are diagonal, which means that bits taken from consecutive codewords are cross-spread so that certain bits of the subsequent codeword arrive earlier in the interleaved bit stream than certain other bits of the previous codeword. In the GSM arrangement, the bits taken from a given codeword of a full rate speech channel are spread over a period of 8 bursts such that 57 bits of the codeword go into each burst. Other interleaving patterns are used in the GSM system, depending on the nature of the information to be interleaved (speech, data, access request, etc.).

Continuing with our full-rate speech channel example, the bursting part of the interleaving / bursting element 104 adds 57 bits of a particular B code word to odd bit positions in the burst and 57 bits (B + west codeword to an interleaved bit bit space, so to get so called. Further, it adds three zero bits (edge ​​bits) to the beginning and end of the burst, and a so-called training sequence of 26 bits exactly in the middle of the burst. The data stream resulting from the output of the interleaving / bursting element thus consists of bursts having a total of 148 bits each. Thus, the description in this patent application would be consistent throughout, in the following the bits of a GSM burst are called symbols. In addition, a burst is called a digital burst while it is still in digital form.

Encryption block 105 performs a logical exclusive or operation to prevent unauthorized reception of the transmitted da20 tan between the encoded data symbols of the digital burst and a certain pseudorandom bit sequence. Edge symbols, robbery flag symbols, and a training sequence are not encrypted. After encryption, the digital bursts are fed to modulator / upconverter unit 106, which converts each digital burst into a radio frequency analog oscillation signal amplified in amplifier 107 and fed to antenna 108 for transmission. Because this ana25 logical signal sequence is closely related to a digital burst, it is also called a burst; for the sake of clarity, it may be more precisely called a transfer burst. Several filtering operations take place inside the modulator / upconverter unit 106 and between the Senja antenna 108; the corresponding filter blocks have been omitted from Figure 1 for the sake of graphical clarity. In a GSM TDMA system, each speech 30 channel may use one time slot from a periodically repeated frame of eight consecutive time slots. The transmitter transmits one transfer burst in the busy time slot of each successive frame during an active connection.

The receiver chain required for receiving, demodulating, and decoding the data transmitted by the transmission chain of Figure 1 would include a receiving antenna for receiving a radio35 signal, some filters and amplifiers for amplifying and filtering the received signal, a down converter / demodulator, or ίζ

.. XC equalizer for converting a transmission burst to digital format at baseband, a decryption block for converting the encrypted bits to normal data, a burst / deinterleave block for extracting and deinterleaving data bits, a channel decoder for decoding a channel and a speech decoder /

A D / A converter for converting a decoded digital signal into an analog signal from which the original speech can be reproduced, for example, through a loudspeaker. The operation of the receiver chain blocks is approximately the opposite of the operation of the corresponding blocks of the transmitter chain.

For transmission modes other than full-speed voice transmission, minor changes to the functions of the transmission and reception blocks described above are required. These changes are known to those skilled in the art from ETSI (European Telecommunications

GSM Specifications published by the Standards Institute) and, for example, Michel Mouly, Marie-Bemadette Pautet, "The GSM System for Mobile Communications, Published by Authors, ISBN 2-9507190-0-7, Palaiseau 1992.

The transmission chain of Fig. 1 is also suitable in principle for EDGE transmissions, although the use of higher data rates would require changes in the operation of the blocks. Data requiring a higher data rate would most likely come from a source other than a microphone and a speech encoder, such as a camera and a video encoder *? .. The channel encoder block would work according to the EDGE channel coding scheme, and together with the interleaving / burst shaping element, the encryption block, and the modulator part of the modulator / upconverter unit, it would have to operate much faster than in the basic GSM system. The channel encoder block would also be able to change the amount of channel coding used according to the link matching commands.

The sharpest difference would be due to a different modulation method. In EDGE's 8PSK modulation25 plan, three successive bits of a digital burst generated are mapped to a single transmission symbol. For this reason, in a digital burst, the symbol is said to consist of a group of three consecutive bits rather than a single bit as in GSM. During burst transmission, the transmitter produces transmission symbols at an instantaneous rate of 270 kilos symbols / s, which is the same as in GSM; the difference in efficiency is due to the fact that the 8PSK symbol carries data corresponding to three bits, whereas in GMSK each symbol represents only one bit.

If the radio channel has good propagation conditions, it is possible to interleave two simultaneous connections between the base station and the GSM or EDGE terminals, so that each connection is allowed to use the reserved time slot only in every other frame. This arrangement is called the half-speed traffic channel n

Ζ to reserve. In the allocated time slot, the overall structure of the transmitted burst remains the same, but the data transmission, channel coding and interleaving / burst shaping formulas of the transmitted data must be adapted accordingly. The same approach can be extended to quarter rate traffic channels, where a given connection is allowed to use the reserved time slot only in every fourth frame, and even to eighth rate traffic channels, where a given connection is allowed to use the reserved time slot only in every eight frames.

The known interleaving methods have the disadvantage that they tend to reduce the effective interleaving depth. The effective interleaving10 of the full speed traffic channel is eight frames, the effective interleaving depth of the half rate traffic channel is four frames, etc., and the effective interleaving depth of the octave or octave rate channel is only one frame. Reducing the interleaving depth means increasing the propagation error sensitivity, so that to achieve a certain level of service quality, the signal-to-noise (S / N) or carrier-to-interference (C / I) ratio associated with the connection should increase accordingly. This is not a viable assumption because simulation has shown that, for example, an eighth rate traffic channel would require a C / I ratio of more than 30 decibels. The gravity of the problem is highlighted by the fact that each interlaced connection should have the same excellent propagation conditions at the same time, which is unlikely because the locations of users within a cell can vary considerably.

A straightforward solution to strengthen the system against propagation errors on lower speed traffic channels would be to extend the interleaving scheme depending on the channel speed so that the bits of a given codeword are spread over more bursts. However, this results in such long delays in the interleaving and de-interleaving stages that they are unacceptable.

From U.S. Serial Nos. 60/144307, 60/144491, and 60/144723, which are not publicly known on the Priority Date of the present application, there is known a method and system in which a known burst structure is radically modified so that two successive times can be transmitted in a single time slot. bursts separated by a protection sequence. This solution alleviates the requirement for an excellent good carrier-to-interference ratio, but has the disadvantage that it requires the definition and implementation of an entirely new burst structure, which is why it is not of interest to designers and manufacturers of transceiver equipment.

Λ rt

<img file="FI112992B_D0001.tif" />

It is an object of the present invention to provide a method and arrangement for interleaving multiple connections over time slots of a TDMA system without imposing excessive demands on the carrier / interference ratio. It is a further object of the invention to provide a method and arrangement that allows flexibility in the number of interlaced connections.

The objects of the invention are achieved by allowing two simultaneous connections to use the same time slot and by using coding and decoding arrangements to separate transmissions belonging to different connections.

The transmission method according to the invention comprises the following steps;

- defining a burst structure which consists of symbols and fills one time slot at the radio interface,

filling the burst with symbols, thereby forming a transfer burst, and

- transmitting a transmission burst in one time slot;

and characterized in that the step of filling the burst with symbols further comprises the following sub-steps:

- taking first data symbols and filling in the first part of the burst,

taking other information symbols and filling in the second part of the burst, and

- taking control symbols and filling them with the rest of the burst.

The receiving method according to the invention is characterized in that it comprises the following steps:

- extracting a plurality of first type symbols from the first portion of the received signal in the time slot,

- extracting a plurality of such symbols from a second portion of said received signal during said time slot,

- attempting to decode said first symbols,

attempting to decode said other symbols, and

<img file="FI112992B_D0002.tif" />

- accepting data symbols which have been successfully decoded.

The invention also relates to a transmitter arrangement for maintaining interleaved connections in a multiple user time division division communication system, the arrangement comprising a burst shaping element arranged to form a symbol burst which fills one time slot at the radio interface. It is characterized in that it comprises means for delivering first data symbols and other data symbols to the burst shaping element, such that the burst shaping means is arranged to fill the first part of the transport burst with said first data symbols, the second part of the transport burst with said other data symbols.

The invention further relates to a receiver arrangement comprising a burst decoder, which is sequenced to decode a burst consisting of symbols that fill a radio interface slot into data symbol sequences; characterized in that it includes means for separately decoding the various data symbol sequences extracted from said transport burst by said bursting device and means for accepting the data symbols that have been successfully decoded.

The use of some coding and decoding arrangements to distinguish between two simultaneous ra20 diode transmissions is known, for example, in the field of spread spectrum communication, where each different transmission is spread with a code taken from a group of orthogonal or near-orthogonal spreading codes. The receiver, which recognizes the code of the received signal for which it is inverted, performs an assembly in which the orthogonality or near orthogonality of the spreading codes cancels out the effect of other simultaneous, unwanted signals. However, the application of spread spectrum techniques requires a relatively complex transmitter and receiver device architecture.

According to the present invention, a receiver may well receive up to some of the content of an unwanted transmission if a digital encrypted system ensures that a particular receiver can decode (and interpret) only portions of the received signal containing the information intended to be received by that receiver.

In one embodiment of the invention. According to the downlink embodiment, the digital burst is formatted according to a burst formatting formula, which may well be the same as that used in any other prior art communication system, but in which the symbols or data bits forming the digital burst come from at least two differently encoded (or The receiver receiving such a burst builds a plurality of code words from the received data bits according to a given interleaving pattern. The receiver is capable of successfully decoding (or interpreting or arranging) only those codewords encoded (or encrypted or scrambled) in the transmitter by the code associated with that receiver.

In one embodiment of the invention. according to the uplink application, the various transmitters may transmit 10 simultaneously, so that only a portion of the payload portion of the transmission burst transmitted by each transmitter consists of actual data with the other areas of the payload portion empty. The sharing of payload parts is coordinated between the transmitters so that in the time dimension only one transmitter at a time transmits the actual data. The transmitters use their own training cycles in the transmission burst so that the receiver can use the training cycles to form a channel estimate for each propagation channel between the transmitter and itself. The training sequences from different transmitters can be concurrent in time, whereby the receiver uses a technique called co-channel estimation to distinguish the training sequences from different transmitters.

Flexibility can be introduced in the method and system by introducing new traffic channel coding schemes that reduce gaps between known full-rate, half-rate, quarter-rate, and eighth-rate formulas. These coding formulas, together with the bursting applications described above, provide a wide variety of interleaving configurations.

The novel features which are considered to be characteristic of the invention are set forth in particular in the appended claims. However, the invention itself, in terms of its structure and mode of operation, as well as its other objects and advantages, will be best understood from the following description of certain embodiments, with reference to the accompanying drawings.

Figure 1 illustrates a known transmission chain in a transmitter,

Figure 2 illustrates the structure and operating principle of an arrangement according to a first embodiment of the invention,

Figure 3 shows an example of a transfer burst in the arrangement of Figure 2,

Figure 4 shows an alternative transfer burst in the arrangement of Figure 2,

Figure 5 shows an overview of the arrangement of Figure 2,

Figure 6 shows an example of a transfer burst in the arrangement of Figure 5,

Figure 7 illustrates the structure and operation of an arrangement according to another embodiment of the invention,

Figure 8 shows examples of a transfer burst in the arrangement of Figure 7,

Figure 9 shows an overview of the arrangement of Figure 7,

Figure 10 shows examples of a transfer burst in the arrangement of Figure 9, and Figure 11 shows a further developed version of the arrangement of Figure 7.

Figure 1 was described in connection with the description of the prior art.

Figure 2 illustrates an arrangement in which a base station subsystem 201 (or generally a transmission arrangement) is in communication with a mobile station 202 (or generally a receive 10 arrangement). The base station subsystem includes a first transmission subchain, which is a serial connection of a first data source 210, a first parity coding unit 211, a first editor or encryption unit 212, a first channel encoder 213, and a first interleaved 214. Connected thereto, the base station includes a second transmission subchain, which is a serial connection of a second data source 220, a second priority encoding unit 221, a second decoder or encryption unit 222, a second channel encoder 223, and a second interleaved 224. The outputs of each of the interleaved 214 and 224 are connected to a burst shaping unit 230, the output of which is connected via a serial connection of modulator 231 and transmitter 232 to a transmitting antenna 233.

Mobile station 202 includes a receiver antenna 240 and, connected thereto, a serial connection of receiver 241, demodulator 242 and burst 243. The output of burst 243 is, in principle, coupled to the first receive subchain and the second receive subchain, although in practice the arrangement may have only one receive chain apparatus whose operation is time-shared. The first receive sub-chain consists of a series connection of a first deinterleaver 250, a first channel decoder 251, a first counter decoder unit 252, and a first parity check unit 253. The second receive sub-chain consists of a sar30 connection of a first deinterleaver 260, a second channel decoder 251, a second receiver decoder unit 252, and a second parity checker unit 263. The outputs of each of the parity check units 253 and 263 are coupled to a data sink block 270.

The arrangement of Figure 2 works as follows. Each of the data sources 210 and 220 produces data to be transmitted to the mobile station 202. Typical data sources in the base station system are logical connections to the public mobile network which, for example, become digitally coded speech to be transmitted to the mobile station. The data consists of

Γ, of bits, and is organized into a number of discrete units that can be called data blocks. The first parity coding unit 211 calculates a parity code for each digital data block from the first data source 210 and adds it to the data block. The parity-coded data block is then scrambled (or encrypted) by the first scrambling or encryption unit 212 using the first scrambling or scrambling code. In the following, only shuffling is discussed briefly. The parity-coded and mixed data block is channel-coded in a first channel decoder 213, whereby bits are mapped into symbols and codewords are formed, and symbols taken from a plurality of consecutive codewords are interleaved in a first interleaver 214. different code for data mixing.

Burst shaping unit 230 takes interleaved symbols from both interleavers 214 and 224 as components of common digital bursts. In order to maintain compatibility with tuned net cellular radio systems, it is preferable to use a known format in digital bursts. Figure 3 shows an example of a digital burst corresponding to known GSM and EDGE specifications. The digital burst, in this order, consists of three leading edge symbols 301, a first data field 302 having 57 data symbols, a first robbery ticket symbol 303, a 26 check 20 training symbol 304, a second robbery ticket symbol 305, a 57 data symbol 306, and a second data symbol 307. The difference with the known GSM or EDGE arrangement is that the data symbols in the first data field 302 come from the output of the first interleaved 214 and the data symbols in the second data field 306 come from the output of the second interleaved 224. Also, each robbery ticket symbol is associated with only one data field. In the picture, the shading emphasizes that certain fields are associated with certain connections. The modulator 231 converts the digital burst to a transmission burst according to known methods, and is transmitted by the transmitter 232 via the transmit antenna 233.

When the mobile station 202 receives a transmission burst from the receiving antenna 240, it converts it into baseband symbols using receiver 241 and a demodulator according to known methods. The burst 243 directs the symbols to the deinterleaver 250 and 260 as it would if the mobile station only received conventional transmission bursts. The data symbols obtained from the first data field 302 go to the first deinterleaver 250 and the data symbols obtained from the second data field 306 go to the second deinterleaver 260. The same procedure is repeated for each transport burst, so that for each transport burst

S, .-:

Π ^r)

- A £.

the interleaved data symbols from the first data field accumulate in the first deinterleaver 250 and the interleaved data symbols in the second data field of each transmission burst accumulate in the second deinterleaver 260. The deinterleaver 250 and 260 invert the interleaving effect to form the original code words. The code words are channel decoded in the respective channel decoders 251 and 261, and despreading of the results is attempted in the descrambler units 252 and 262.

It can now be assumed that the first data source 210 of the base station subsystem 201 produced data to be transmitted to the mobile station 202, while the second data source 220 of the base station subsystem 201 produced data to be transmitted to another mobile station. The mobile station 202 knows the code to be used in the first encoder or encryption unit 212 because the code was agreed upon when establishing a communication link between the base station subsystem 201 and the mobile station 202. The mobile station does not know the code used by the second editor or encryption unit 222. Thus, the reordering operation in the first reordering unit 252 succeeds because the mobile station uses the correct reordering code, but the reordering operation attempted in the second reordering unit 262 fails because the mobile station uses the wrong reordering code. The success and failure of the organizing operations performed in the countermeasure units 252 and 262 is noted in the respective parity checker units 253 and 263, since only a properly organized data block corresponds correctly to the associated parity code. Only the data block whose parity code matched the data block is input to data sink block 270.

Alternatively, we can assume that both data sources 210 and 220 in the base station subsystem 201 provided data to be transmitted to mat25 mobile station 202, but for different channel allocation or other reasons, two different scrambling codes were used. If the respective two reordering codes were known to the mobile station 202, it would be able to make one of them available for use in the first transducer unit 252 and the second in the second transducer unit 262, so that data can be organized in both units and data blocks

253 and 263 found in the parity checks performed, is input to the data sink block 270.

The invention does not place any restrictions on the method of compressing symbols into bursts in a base station subsystem and respectively extracting bursts in a mobile station. Figure 4 illustrates an alternative digital burst configuration in which data symbols 302 and 306 alternate symbols associated with different connections

ζ.

symbol by principle. The mobile burst is responsible for separating symbols for correctly selected deinterleavers.

The arrangement of Figure 2 can be easily generalized to include an arbitrary number of interleaved communication connections. Figure 5 illustrates an arrangement in which base station subsystem 501 comprises a plurality of parallel transmission subchains illustrating chains 510 through 530. The burst shaping element 540 takes symbols from the output of each transmission subchain when assembling digital bursts. Conceptually, the mobile station 502 comprises an equal number of receiving subchains connected to the bursting unit 550, of which the receiving subchains 560 to 580 are shown. Again, in practice, conceptually parallel units may be implemented in the same device. Figure 6 illustrates a burst structure having N / 2 data fields between the leading edge symbols 601 and the training sequence 602 and an N / 2 data field between the training edge 602 and the trailing edge symbols 603. Each data field is shown to have its own robbery flag symbol S. Here, N is a positive integer.

The advantage of the arrangements of Figures 2 and 5 over the prior art arrangements is that each interleaving unit can be programmed to interleave data symbols from a codeword from an arbitrary set of digital bursts. For example, the interleaving depth of eight frames can be maintained by interleaving symbols from a codeword from eight digital bursts. Of course, all units of deinterleaving and de-interleaving20 must be aware of the interleaving formula used.

Next, the so-called "invention" of the invention will be described. 7 to 10. Referring to FIGS. 7 to 10, FIG. 7 illustrates a first mobile station 701 and a second mobile station 702 (or, generally, first and second transmission arrangements), each communicating with a base station subsystem 703 (or, generally, a receiving arrangement).

The transmission chain in each of the mobile stations is similar to a serial connection comprising a data source 710 (720), a parity coding unit 711 (721), an optional modifier 712 (722), a channel coder 713 (723), an interleaver 714 (724), a burst shaping element 715 (725), 726), transmitter 717 (727) and transmit antenna 718 (728). In the base station subsystem, the receiving antenna 730 is coupled, via the receiver 731, to a common channel estimation element 732 which communicates with two parallel receiving share chains. Each receive subchain is a serial connection comprising an equalizer 740 (750), a burst 741 (751), a deinterleaver 742 (752), a channel decoder 743 (753), an optional counter modifier 744 (754), a parity checker 745 (755) and a dataan ).

Λ Γ- ·, ί -, · '3 /

The arrangement of Figure 7 operates as follows. Each of the data sources 710 and 720 produces data to be transmitted to the base station subsystem 703. Typical data sources in the mobile subsystem are source data codecs that become e.g.

digitally encoded speech to be transmitted to the public mobile network. The data consists of bits and is organized into certain discrete units which can be called data blocks. Parity coding units 711 and 721 compute the parity code for each digital data block from data sources 710 and 720 and add the parity codes to the data blocks. The parity-coded data blocks are then mixed (or encrypted) in the editor or encryption units 712 and 722, if mixing is deemed necessary; each mobile station uses its own scrambling or encryption code. The parity-encoded and mixed data blocks are channel-coded in the channel encoders 713 and 723, mapping bits into symbols and forming code words, and symbols taken from a plurality of consecutive codewords are interleaved in interleavers 714 and 724.

Burst shaping units 715 and 725 receive interleaved symbols from their respective interleavers 714 and 724 as components of digital bursts. Again, in order to maintain compatibility with known radio systems, it is advantageous to use the known format in digital bursts only in a slightly modified manner. Figure 8 is an example of a pair of digital bursts that generally correspond to known GSM and EDGE specifications. The first digital burst 800 consists of three leading edge symbols 801, 57 (first) data fields 802, a (first) robbery ticket symbol 803, a 26 check symbol training sequence 804, a false capture ticket symbol 805, 57 a pseudo symbol sequence 806, and a fake symbol field 806. The second digital burst 810 consists of three leading edge symbols

811, False Data Field 812, 57 False Symbol Flags

813, 26 training symbols 814, a check symbol 816, 57 data symbols 816, 57 data symbols 816, and three trailing edge symbols 817, respectively. In other words, a change from the known GSM and EDGE specifications is that the second plunder flag symbol and second data field of the first digital burst 800 and the first data field and first plunder flag symbol of the second digital burst 810 are replaced by pseudo symbols. In addition, each mobile station uses its own training sequence in fields 804 and 814.

False symbols can be some predefined standard symbols, or they can denote moments where there is no symbol transmission (cf. the edge symbols which indicate that no symbol transmission occurs). It is probably best to use the latter option, because if nothing is transmitted during the Vale symbols, this means less interference and unnecessary transmission power. Keeping the overall interference level low is advantageous because it directly improves the total available capacity of the cellular radio system. Transmit power saving is especially important in portable radio terminals where it helps to extend battery life between two consecutive charges. For the sake of consistency, throughout this patent application we will use the term pseudo-symbols also in connection with embodiments of the invention in which nothing is actually transmitted about the pseudorandom of the transport burst.

In Figure 8, the shading highlights non-false fields associated with certain connections. In the arrangement of FIG. 7, modulator 716 (726) converts the digital burst to a transmission burst according to known methods, and is transmitted by transmitter 717 (727) via transmit antenna 718 (728).

When the base station subsystem receives simultaneously transmitted transmission bursts originating from digital bursts 800 and 810 via a receiving antenna 730 and a receiver 731, it obtains channel estimates from these two different training sequences 804 and 814 with the common channel estimation means 732. that described in PA Beach, A. Hottinen and Z.-C. Honkasalo: “Co-channe!

Interference Canceling Receiver for TDMA Mobile Systems, "Proc. ICC'95, pp. 17-21, 1995, which is incorporated herein by reference. Once the base station subsystem has received channel estimates for the propagation channels between it and the two mobile stations, it is able to successfully equalize and decode the data fields from both transport bursts at the parallel equalizers 740 and 750 and the parallel decoders 743 and 753. Note that the use of pseudo-symbols ensures that the base station subsystem does not receive overlapping data symbols over time.

Receiving successive transfer bursts at the respective time slots and co-channel estimation and sorting the dataybols to the equalizers 740 and 750 results in the data symbols from the first mobile decoder 701 accumulating in the first decompressor 742 and the second mobile decoder 752. Deinterleaving again reverses the interleaving effect in both reception; chains, followed by simple channel decoding, organization (if necessary), and parity checking in the respective policy areas.

Γ.

Ζ

The arrangement of Figure 7 can be generalized to include an arbitrary number of interleaved communication connections. FIG. 9 illustrates an arrangement of a plurality of parallel mobile stations, showing mobile stations 901-903 in the figure. Base station subsystem 904 comprises a plurality of parallel receive sub-chains, of which the reception 5 sub-chains 910-930 are shown. The co-channel estimation body 940 serves all reception share chains. Figure 10 illustrates a plurality of bursts of which bursts 1001 1003 are shown in the figure: there are N / 2 data fields between the leading edge symbols 1011 and the training sequence 1012, and N / 2 data fields between the training sequence 1012 and the trailing edge symbols 1013. Each field is shown to have its own robbery flag symbol S, and N is a positive integer. Each frame uses only one data field, its associated robbery flag symbol and training sequence, and the other data fields and robbery · flag symbols are fake fields or symbols.

A patent application filed simultaneously on the priority date of this application by the same applicant discloses a method and arrangement for repeatedly improving channel state. The solution presented therein is based on the channel estimate being repeatedly updated with information from channel decoding. Such a solution is particularly well suited for improving co-channel expression in a base station subsystem. Fig. 11 is a developed version of the communication system of Fig. 7, wherein feedback for repeated channel estimation and equalization is sequenced through re-interleaving blocks 1147 and 1157.

The exemplary embodiments of the invention described above are not to be construed as limiting the scope of the appended claims. In particular, certain additional features shown as further developments of the basic invention are not mutually exclusive, but can be combined in many different ways.