RU2208911C2 - Method of diversified signal transmission and device for its realization - Google Patents

Abstract

FIELD: radio engineering. SUBSTANCE: method consists in formation of N diversified signal transmission channels. From flow of symbols there are formed packages of M > N symbols if N is not equal to degree of number two. N orthogonal sequences of symbols are so formed from symbols of each package that each of them has all symbols of package. Principle of formation of orthogonal sequences of symbols should be known at receiving side. Diversity channel of each formed orthogonal sequence of symbols in each package is assigned and consistent transmission of package is executed. Device for realization of method includes former of packages of symbols, former of pilot signals, N modulators, N transmission antennas, unit computing length of package of symbols and former of orthogonal sequences of symbols. EFFECT: raised noise immunity of radio communication systems. 9 cl, 13 dwg

Description

 The invention relates to the field of radio engineering, in particular to a method of diversity transmission of a signal and a device for its implementation.

 One of the main reasons for reducing noise immunity and, as a consequence, throughput in mobile radio communications is signal fading. To combat them, various types of diversity are used, the most radical of which is a spatially-spaced technique. For a number of technical and economic reasons, spatially distributed reception is mainly used at base stations (BS), i.e. in the reverse channel of radio communication systems. In the forward channel (communication direction from the BS to the mobile station (MS)), time and / or frequency diversity is typically used.

 The disadvantage of these two types of diversity is that they require transmission time or frequency band to duplicate the transmitted information. The loss of efficiency resulting from such duplication can be avoided by combining diversity with other processing methods. For example, a variation of temporal diversity is the interleaving of coded symbols [1, J. G. Proakis Digital Communications, NY, 1995] in a transmitter. In communication systems with SHPS, frequency diversity is realized due to the wide frequency band occupied by the signal. However, in this case, there are significant disadvantages.

 Since interleaving provides diversity not at the symbolic level, but at the level of the packet of encoded symbols, it is only able to mitigate to some extent the effect of fading on the noise immunity of reception. In this case, discontinuities in the complex envelope of the signal after deinterleaving make it impossible for optimal joint demodulation and decoding of the received signal. Another significant drawback of temporal diversity is that, due to the restriction on the delay in transmitting information, temporal diversity becomes ineffective at low MS speeds when the correlation interval of the complex envelope of the signal becomes comparable or greater than the interleaving interval.

 The implementation of frequency diversity in communication systems with noise-like signals (SHPS) has a negative side effect. On the one hand, the frequency diversity efficiency increases with an increase in the width of the NPS spectrum, while in the time domain it becomes possible to distinguish and independently process the multipath components (rays) of a signal whose relative time delays exceed the peak width of the NLS autocorrelation function. But at the same time, the signal-to-noise ratio in each ray decreases, since the rays are mutual additive interference.

 In the direct channel of CDMA systems, orthogonal codes are used to separate user channels [2, Andrew J. Viterbi. CDMA Principles of Spread Spectrum Communication. / ADDISON-WESLEY PUBLISHING COMPANY, April 1995.], which allows to increase the noise immunity of information transmission in a stationary channel and in a channel with flat fading. But with frequency-selective fading, the effect of multi-user interference occurs, which consists in the fact that the entire group BS signal in the remaining resolvable beams is an obstacle to the demodulation of each beam of the user signal.

 The indicated shortcomings of the time and frequency diversity make the task of developing the diversity method in the forward channel, comparable in efficiency with spatially diversity reception in the reverse channel, relevant.

 Until recently, of all the separated BS antennas, only one was used for transmission. Meanwhile, it is obvious that the use of diversity transmitting antennas has a significant potential for increasing the noise immunity of information transmission. For example, if the complex transmission coefficients of the signals between the BS antennas and the MS antenna are known, then this information can be used to optimally set the amplitude and phase of the transmitted signal in each antenna. In this case, due to the transmission of the signal from the diversity antennas, the same gain in noise immunity of reception in the forward channel will be ensured as those of the same antennas for spatially diversity reception in the reverse channel [3, Microwave Mobile Communications / Edited by William C. Jakes. IEEE Press. NY, 1994].

 However, due to obvious problems in obtaining information about the communication channel, the described diversity transmission algorithm is difficult to implement in practice. In this regard, of great practical interest is the development of a method of diversity transmission and a device for its implementation, which does not require information about the communication channel for its implementation.

Various methods and devices for diversity signal transmission are currently known. Diversity signal transmission is carried out from two or more space-separated antennas, as shown in figure 1, while the implementation of these methods does not require information about the communication channel. These methods of diversity transmission include the following methods:
CTD (coded transmit diversity) - code diversity transmission, [4, European Patent Application number 0 605 119, Diversity for direct-sequence spread spectrum systems. Priority: 12/29/92. Int. Cl. ⁵ H 04 B 7/06;
DTD (delayed transmit diversity) - delayed diversity transmission [5, A. Witneben "Base Station Modulation Diversity for Digital SIMULCAST," Proceeding of the 1991 IEEE Vehicular Technology Conference (VTC 41 st), pp. 848-853, May 1991], [6, A. Witneben "A New Bandwidth Efficient Transmit Antenna Modulation Diversity Scheme For Linear Digital Modulation," in Proceeding of the 1993 IEEE International Conference on Communications (IICC'93), pp. 1630-1634, May 1993];
OTD (orthogonal transmit diversity) - orthogonal diversity transmission [7, Draft UMTS-2000 standard for cellular systems with code division multiplexing, developed by ETSI-SMG2. Submission of Proposed Radio Transmission Technologies, and published January 29, 1998, pp. 51-52, section 5.6.3.1 Orthogonal Transmit Diversity];
PSTD (phase sweeping transmit diversity) - Phased rotation transmission [8, A. Hiroike, F. Adachi and N. Nakajima, "Combined effects of phase sweeping transmitter diversity and channel coding," IEEE Trans. Veh. Technol., Vol. 41, pp. 170-176, May 1992];
STTD (space-time transmit diversity) - space-time diversity transmission [9, WO 99/14871 of 03/25/99, "Transmitter Diversity Technique For Wireless Communications", Int. Cl. ⁶ H 04 B 7/06, 7/08].

In describing the diversity transmission method, the term “symbol” will be used hereinafter. Let us explain the meaning of this concept. Any informational message can be recorded as a sequence of characters belonging to a certain alphabet. For the most commonly used bit representation of information, the information message is presented as a sequence of binary characters (bits), each of which can take the value 0 or 1. For the purpose of increasing the efficiency of data transfer in communication systems, other alphabets are used except binary. In this case, the original bitstream is divided into blocks of m bits. Moreover, M = 2 ^m possible combinations of the values of the bits forming each block are the symbols of the new alphabet Ω.

где A_sm - представляет собой амплитуду синфазной несущей частоты (cosω₀t), а A_sm - амплитуду квадратурной несущей частоты (sinω₀t) на интервале текущего символа. Эту величину иногда называют символом модуляции. Заявляемое изобретение (описание и формула) оперирует именно с символами модуляции. Поскольку каждому символу однозначно соответствует символ модуляции, при описании изобретения и в формуле изобретения второе слово в термине "символ модуляции" опускается и используется более короткий термин "символ".Information is transmitted through the communication channel by modulating the parameters of the transmitted carrier oscillation. In the most common case of amplitude-phase modulation (MPSK, MQAM modulation types), each message symbol is associated with a real or complex quantity a _m = A _cm + jA _sm ,

where A _sm - is the amplitude of the in-phase carrier frequency (cosω ₀ t), and A _sm is the amplitude of the quadrature carrier frequency (sinω ₀ t) in the interval of the current symbol. This value is sometimes called the modulation symbol. The claimed invention (description and formula) operates specifically with modulation symbols. Since each symbol is uniquely associated with a modulation symbol, in the description of the invention and in the claims, the second word in the term “modulation symbol” is omitted and the shorter term “symbol” is used.

 The CTD code diversity transmission method [4] consists in modulating it with its orthogonal sequence before transmitting each message symbol in each antenna. For example, Walsh functions can be used as such sequences. Due to the mutual orthogonality, the symbol copies transmitted by different antennas can be separated by the MS receiver and then optimally summed, for example, in accordance with the MRC (maximal ratio combining) algorithm [10, William C. Lee. Technique of mobile communication systems. Per. from English V.N. Talyzina / Ed. THEM. Pyshkina. M .: Radio and Communication, 1985].

 In the absence of propagation channel information, the CTD method provides the greatest noise immunity for diversity transmission. However, CTD has a significant drawback, since orthogonal modulation of the symbol in the transmitting antennas leads to an expansion of the spectrum of transmitted signals by N times, where N is the number of transmitting antennas.

 The method of delayed diversity transmission DTD [5, 6] consists in converting flat fading in the communication channel into frequency selective and using the received frequency diversity when receiving. To do this, the same signal is transmitted through different antennas, but with relative delays. The delay in signal transmission between any two BS antennas should be no less than the reciprocal of the width of the spectrum of the transmitted signal. In CDMA systems, this value is equal to the duration of the spread spectrum chip. As a result, the signal at the receiving point is no different from the usual multipath signal, and therefore its processing is carried out by a standard multipath (Rake) receiver [1]. The advantage of DTD is that its implementation does not require changes to existing communication standards. The payment for the frequency diversity implemented in the method is intersymbol interference in communication systems using simple signals, and multi-user interference in CDMA systems. Let us explain the latter in more detail. In the direct channel of CDMA systems, orthogonal codes are used to separate user channels, which makes it possible to increase the noise immunity of information transmission under the conditions of a stationary channel and in a communication channel with flat fading. But with frequency-selective fading, the effect of multi-user interference occurs, which consists in the fact that the entire group BS signal in the remaining resolvable beams is an obstacle to the demodulation of each beam of the user signal.

 The method of diversity transmission with rotation of the phase PSTD [7] allows you to increase the noise immunity of the transmission of encoded information in the channel with fading by increasing the efficiency of interleaving. Due to the limitation on the data transmission delay, the interleave interval, as a rule, is a relatively small amount (10-20 ms). Therefore, temporary interleaving is effective only at a sufficiently high frequency of fading. The latter is proportional to the speed of movement of the subscriber and is often low, for example, when communicating with stationary or slowly moving subscribers. The idea behind PSTD is to increase interleaving efficiency by artificially increasing the frequency of fading. To do this, the same signal is transmitted through two spaced antennas, but in one of the antennas the signal is additionally modulated according to some law F (t). Reception is carried out in the usual way, as in the case of signal transmission through only one antenna.

физический смысл которой заключается в сдвиге частоты несущего колебания в одной из антенн на величину f_sweep. Частота искусственных замираний при этом также равна f_sweep.The presence of additional modulation causes a continuous change in the range [0.2π] of the phase difference of the signals coming from different antennas. Due to the interference of these signals at the receiving point, a fading effect is created, the frequency of which is determined by the function F (t). As the last in [7], it is proposed to use a function of the form

the physical meaning of which is to shift the frequency of the carrier wave in one of the antennas by f _sweep . The frequency of artificial fading is also equal to f _sweep .

 PSTD, like DTD, does not require changes to existing wireless standards. The main disadvantage of PSTD is that it explodes at the frame level of coded symbols, which is less efficient than diversity at the level of each symbol. The disadvantages are the deterioration of the operating conditions of the carrier recovery scheme due to an increase in the frequency of fading, and the lack of the possibility of using more than two transmitting antennas in the method.

 The method of orthogonal diversity transmission OTD [8] is that the input stream of encoded symbols is divided into N sub-streams, which are transmitted through different antennas. Different orthogonal codes, for example, Walsh codes, are used to modulate the symbols in the antennas. Due to this, the symbols transmitted by different BS antennas are separated when received at the MS. Since the duration of the symbol in each substream is N times longer than the duration of the symbol of the original stream, additional modulation with orthogonal codes does not expand the spectra of the emitted signals relative to conventional signal transmission through one transmit antenna.

 To implement coherent reception, a unique orthogonal pilot signal must be transmitted through each antenna, which evaluates the complex envelope of the signal of the corresponding BS transmitting antenna in the MS receiver.

 Since each symbol is transmitted through only one antenna, OTD does not provide symbol diversity. Instead, the OTD accepts encoded packet symbols with independent fading, i.e. performs the function of an interleaver, but not in the temporal, but in the spatial dimension. The effectiveness of this spatial interleaving does not depend on the frequency of fading and is the higher, the greater the number of transmitting antennas used.

The method of orthogonal diversity transmit-receive signal in a cellular radio communication system with code division multiplexing [11, RF patent 2145152, IPC ⁶ N 04 V 7/216] is that on the transmitting side:
each information stream of binary characters of each user is assigned an expanding orthogonal code,
form N spatially separated transmission channels,
generate a pilot signal for each spatially separated transmission channel, and all pilot signals and expanding user codes are mutually orthogonal,
each binary symbol information stream is divided into sequential information packets of N sequentially located binary symbols in each,
carry out serial-parallel conversion of binary characters in each serial information packet, forming a parallel information packet of N binary symbols,
repeating the information parallel packet N times, thus forming an information serial-parallel packet that contains N parallel and N consecutive groups of binary symbols on the interval of the duration of the serial information packet, for each sequential group of information serial-parallel binary symbols,
forming an orthogonal code of N binary symbols, thus forming a series-parallel binary symbol package of orthogonal codes, which contains N parallel and N consecutive groups of binary symbols of orthogonal codes,
rearrange binary symbols in parallel groups of an information serial-parallel packet so that binary symbols of each sequential group are not repeated,
rearrange binary symbols in a serial-parallel binary symbol packet of orthogonal codes, the same as in the information serial-parallel binary symbol packet,
summing modulo two each binary symbol of the information serial-parallel packet of binary symbols with the corresponding binary symbol of the serial-parallel packet of binary symbols of orthogonal codes, thus forming a series-parallel packet of binary encoded symbols,
carry out the expansion of binary coded symbols of a serial-parallel packet by summing modulo two of each binary encoded symbol with an orthogonal code assigned to the user information stream, thereby forming an information serial-parallel packet of extended binary symbols,
a pilot signal for each spatially separated transmission channel is generated with a repetition period that is a multiple of the duration of the packet of binary symbols,
assign each spatially distributed channel to each successive group of extended binary symbols of the information serial-parallel packet,
modulate and transmit through N spatially separated transmission channels simultaneously in time successive groups of extended binary symbols of each information stream of each user and the corresponding pilot signals.

 A device for implementing the method of orthogonal diversity transmission-reception of a signal in a cellular radio communication system with code division multiplexing according to patent 2145152 [11] contains, for example, for M users M of similarly formed signal transmission branches. Each signal transmission branch contains the following blocks.

 A shaper of a serial-parallel binary symbol packet, in which the binary symbol information stream is divided into serial information packets, binary-serial conversion of binary symbols is performed, forming a parallel packet of N binary symbols, the parallel information packet is repeated N times, forming an information serial-parallel packet of N binary characters.

 A binary symbol permutation unit that performs binary symbol permutation in parallel groups of an information serial-parallel packet.

 Block forming a serial-parallel package of binary symbols of orthogonal codes.

 The binary symbol permutation block of the orthogonal codes, which carries out the binary symbol permutation of the orthogonal codes, is the same as in the information serial-parallel binary symbol packet.

 The first modulo-two summing unit, in which modulo two each binary symbol of the information serial-parallel packet of binary symbols is summarized with the corresponding binary symbol of the serial-parallel packet of binary symbols of orthogonal codes, thus forming a series-parallel binary encoded symbol packet.

 A spreading orthogonal code generator that generates a spreading orthogonal code of a user information stream.

 The second modulo-two summing unit, in which the binary coded symbols of the serial-parallel packet are expanded by modulo-summing two of each binary coded symbol with the orthogonal code assigned to the user information stream, thus forming an information serial-parallel packet of extended binary symbols.

 A block for generating pilot signals, N modulators and N antennas.

 The pilot signal generation unit generates a pilot signal for each spatially separated transmission channel with a repetition period that is a multiple of the duration of the binary symbol packet, and all pilot signals and user extension codes are mutually orthogonal.

 Modulators in which modulate successive groups of extended binary symbols of each information stream of each user and the corresponding pilot signals.

 Antennas for transmitting successive groups of extended binary symbols of each information stream of each user and the corresponding pilot signals.

 Thus, in the analogue method and device for its implementation, each binary symbol of the transmitted stream is transmitted through N diversity antennas, and due to special transformations to which the binary symbols are subjected, the transmitted binary symbols are orthogonal to the receiving point. The described technical solution allows the use of an arbitrary number of transmitting antennas. Due to the fact that each binary symbol of the transmitted symbol stream is transmitted through all transmit antennas and the orthogonality of the transmitted binary symbols at the receiving point is ensured, this method allows to obtain the greatest noise immunity of the diversity transmission.

 However, the orthogonality of the transmitted symbols at the reception point in the described analogue method [11] is provided only when the binary symbols are real numbers. In the case of complex symbols, orthogonality is not ensured, and therefore, binary symbols of a transmitted packet may cause mutual interference upon reception. Orthogonality violation can also occur when the number of transmitting antennas is not equal to the power of two, since in this case it is impossible to provide orthogonality of strings of a parallel-parallel packet of encoded symbols in this method. In these cases, the noise immunity of the diversity transmission will decrease.

The closest technical solution to the claimed invention is a method of space-time diversity transmission of STTD [9, WO 99/14871 dated 03/25/99, "Transmitter Diversity Technique For Wireless Communications", Int. Cl. ⁶ H 04 B 7/06, 7/08], which consists in the fact that the input symbol stream is divided into packets of M symbols, each symbol packet is encoded, as a result of which M • L channel symbols are formed, N transmit antennas are formed, channel transmit symbols through N transmitting antennas during K symbol intervals, distributing channel symbols across N transmitting antennas during K symbol intervals in such a way that either N is M (N = M) or K is L (K = L); either N is L (N = L), or K is M (K = M), and a channel symbol corresponding to each symbol in the input stream is transmitted through each transmit antenna. An embodiment of this method for the preferred case N is 2 (N = 2) transmitting antennas.

 The description of the invention is fully devoted to the preferred case of N = 2. In particular, various processing options for the transmitted signal at the receiving side are considered.

 In the described invention, a device for implementing the prototype method is shown (shown in FIGS. 1 and 2 of the patent description). A device that implements the proposed sequence of operations on the transmitting side is presented in a very generalized form, i.e. contains an encoder and N transmit antennas. The operation algorithm of this device is also described in a generalized way by listing the actions by which the solution of the task is achieved. Therefore, it is not clear from the description and claims how this invention is implemented.

 The main disadvantage of this invention is that it has low efficiency for the case of N greater than 2 (N> 2) diversity channels (for example, diversity transmit antennas). The claims refer to a certain encoding algorithm for a packet of M symbols and indicate that the channel symbols obtained as a result of this encoding algorithm are distributed among N transmitting antennas during K symbol intervals (different frequencies). However, neither the encoding algorithm itself, nor at least the principles that underlie it are disclosed either in the formula or in the description. It does not say how the number of characters in the packet and the parameter L are related. Analysis of the invention for N> 2 transmitting antennas shows that it is impossible to eliminate the mutual interference of the symbols of the transmitted packet using the operations and relations between the parameters M, L, N, and K at reception. The presence of these mutual interference will lead to a significant reduction in noise immunity.

 The task of the invention is to increase the noise immunity of radio communication systems due to the diversity of signal transmission in the channel with fading.

 The problem is solved as follows.

In the method of diversity transmission of the signal, which consists in forming N diversity channels of signal transmission, packets of N symbols are formed from the input symbol stream, according to the invention, a new sequence of actions is introduced:
form packets of M greater than N characters, if N is not equal to the power of two,
from the symbols of each packet, N orthogonal symbol sequences are formed in such a way that each of them contains all the symbols of the packet, and the rule for generating orthogonal symbol sequences must be known on the receiving side,
assigning a diversity channel to each generated orthogonal symbol sequence of each packet and sequentially transmitting the packets.

 The input symbol stream contains information and pilot symbols.

 Packets of M characters can be formed in various ways, therefore, the dependent claims provide examples of how this can be done in practice.

 Packets of M symbols are formed from the input symbol stream so that each packet consists of information symbols only or only pilot symbols.

 Forming packets of M symbols, M is chosen to be greater than or equal to N if the number of diversity channels is equal to a power of two; in the opposite case, M is chosen to be the nearest number N that is equal to a power of two.

 Forming N orthogonal symbol sequences, N different dyadic shifts of the original symbol sequence of the packet are performed, forming N symbol sequences.

 If the symbols of the packet are real numbers, then, forming N orthogonal sequences, the symbols in the formed sequences are multiplied by coefficients equal to + A or -A, where A is a non-zero real number, so as to ensure the orthogonality of all the generated sequences, regardless package symbol values.

 If the symbols of the packet are complex numbers and N = 2, then, forming orthogonal sequences, multiply the symbols in the obtained sequences by coefficients equal to + A or -A, where A is a non-zero real number, and change the values of some of the symbols to complex their conjugate values in such a way as to ensure the orthogonality of all generated sequences, regardless of the values of the packet symbols.

 If the symbols of the packet are complex numbers and N> 2, then, forming N orthogonal symbol sequences, each formed sequence of symbols is repeated, forming N sequences of 2M symbols long, the symbols in the formed sequences are multiplied by coefficients equal to + A or -A, where A - a non-zero real number, and change the values of part of the characters to complex conjugate values in such a way as to ensure the orthogonality of all the generated sequences independently imo the values of package symbols.

To a diversity signal transmission device comprising a symbol packetizer, a pilot generator, N modulators and N transmitting antennas, the first input of the symbol packetizer is the first input of the device, the first input of each modulator is connected to the corresponding output of the pilot generator , the output of each modulator is connected to its corresponding antenna, according to the invention are additionally introduced
a block for calculating the length of the symbol packet and a block for generating orthogonal symbol sequences, wherein the first inputs of the block for calculating the length of the symbol packet and the block for generating orthogonal symbol sequences are combined to form a second input of the device, the output of the block for calculating the length of the symbol packet is connected to the second inputs of the symbol packetizer and the generating block orthogonal symbol sequences, the third input of which is connected to the output of the symbol packetizer, the fourth input of the block tions orthogonal symbol sequences is a third input device, the second input of each modulator is coupled to its corresponding output of the formation of orthogonal symbol sequences.

 The unit for the formation of orthogonal sequences of characters is given as an example implementation for the claimed device. The unit for generating orthogonal symbol sequences comprises a driver of N dyadic shifts, N parallel branches for generating orthogonal symbol sequences, a storage device, a verification unit and a logical element AND, wherein the first input of the storage device and the input of the verification unit are combined to form the first input of the unit for generating orthogonal symbol sequences, the second input of the storage device is the second input of the block forming the orthogonal sequences of symbols , the first input of the driver of N dyadic shifts is the third input of the block for generating orthogonal symbol sequences, each branch for generating orthogonal symbol sequences contains sequentially connected an orthogonal code superimposing unit, a repetition unit, a first multiplexer, a complex conjugation unit and a second multiplexer, wherein the output of the orthogonal code superimposing unit additionally connected to the second inputs of the first and second multiplexers, the first inputs of the nodes of the overlay orthog channel codes of N branches of forming orthogonal symbol sequences are connected to their corresponding outputs of the driver of N dyadic shifts, the second inputs of nodes of overlapping orthogonal codes of N branches of forming orthogonal sequences of symbols are connected to their corresponding outputs of the storage device, the third inputs of the first multiplexers of N branches of forming branches of orthogonal symbol sequences are combined and connected to the output of the logic element AND, the third inputs of the second multiplexers N branches of the formation of orthogonal symbol sequences are combined with the first input of the logical element And, forming the fourth input of the block forming the orthogonal symbol sequences, the second input of the logic element And is connected to the output of the verification node, the outputs of the second multiplexers N branches of the formation of the orthogonal symbol sequences are outputs of the block forming the orthogonal symbol sequences .

 A comparative analysis of the proposed method of diversity signal transmission with the prototype revealed common and distinctive features according to the claims.

 General features of the proposed method and the prototype method: form N spaced signal transmission channels, from the input stream of characters form packets of N characters.

 Distinctive features of the proposed method from the prototype method: form packets of M greater than N characters, if N is not equal to the power of two, from the characters of each packet form N orthogonal sequences of characters so that each of them contains all the characters of the packet, and the rule of formation of orthogonal symbol sequences should be known on the receiving side, assign the diversity channel of each generated orthogonal symbol sequence of each packet and carry out the sequence th packet transmission.

 The inventive method is developed for the general case of multi-level quadrature (complex) signal modulation and can significantly increase the noise immunity of the radio communication system in the channel with fading.

 Comparison of the proposed method with other known technical solutions in the art [4-9, 11] did not reveal the signs claimed in the characterizing part of the claims.

 A comparative analysis of the inventive device diversity transmission of the signal with the prototype allowed to identify common and distinctive features according to the claims.

 Common features of the claimed device and the prototype device: the device contains a symbol packetizer, a pilot signal generation unit, N modulators and N antennas, as well as the connections listed in the restrictive part of the claims.

 The distinguishing features of the claimed device and the prototype device are as follows: a block for calculating the length of a symbol packet is introduced, which forms packets of M symbols in such a way that M is equal to or greater than N, while M must be equal to a power of two, and a block for generating orthogonal symbol sequences which forms from the symbols of each packet N orthogonal symbol sequences and performs N different dyadic shifts of the original symbol sequence of the packet, forming N symbol sequences in.

 Thus, the claimed device implements in full the features of the proposed method.

 A comparative analysis of the claimed device with other technical solutions [4-9, 11], known in the art, did not allow to reveal the features claimed in the characterizing part of the claims.

 The inventive method and device allows, in comparison with known technical solutions, to significantly increase the reliability of information transfer and the capacity of a radio communication system with code division multiplexing. Therefore, the claimed method of diversity signal transmission and a device for its implementation, created in a single inventive concept, meet the criteria of the invention of "novelty", "technical solution", "significant differences" and meet the inventive step.

 The description of the invention is illustrated by graphic materials.

 Figure 1 illustrates a General view of a method of diversity transmission of a signal in the forward channel of a communication system.

 Figure 2 shows a block diagram of the inventive device diversity transmit signal.

 In FIG. 3 is a block diagram of a block for generating orthogonal symbol sequences, shown as an example of the implementation of this block.

 In FIG. 4 shows an example of packet formation of M symbols from the input symbol stream.

 In FIG. 5 is an example implementation of the proposed method of diversity signal transmission for N = 2, for the case when the symbols are complex numbers.

 In FIG. 6 is an example implementation of the inventive method of diversity signal transmission for N = 4, for the case when the characters are real numbers.

 In FIG. 7 is an example implementation of the proposed method of diversity signal transmission for N = 4, for the case when the symbols are complex numbers.

 On Fig shows the dyadic shifts of the sequence of characters in the package for M = 4, are given as an example of the operation of the dyadic shift.

 Figure 9 shows possible options for replacing part of the characters in N character sequences with complex conjugate values that ensure the orthogonality of these sequences for N = 2.

соответствующие примеру реализации изобретения, показанному на фиг.7.Figure 10 shows the sequence f _m and

corresponding to the example implementation of the invention shown in Fig.7.

 Figure 11 shows the dependences of the probability of erroneous reception of the SNR frame per bit for various methods of diversity transmission obtained by computer simulation at N = 2.

 Figure 12 shows the dependences of the probability of erroneous reception of the SNR frame per bit for various diversity transmission methods obtained by computer simulation at N = 4.

 Figure 13 shows the dependences of the probability of erroneous reception of the SNR frame per bit for various diversity transmission methods obtained by computer simulation at N = 8.

The device (Fig. 2) comprises a symbol packetizer 3, a pilot signal generating unit 4, N modulators 5 ₁ -5 _N and N transmit antennas 6 ₁ -6 _N , the first input of the symbol packetizer 3 being the first input of the device, the first the input of each modulator 5 ₁ -5 _{N is} connected to the corresponding output of the pilot signal generating unit 4, the output of each modulator is connected to its corresponding antenna. According to the invention, the device further comprises a unit for calculating the length of the packet of characters 7 and a unit for generating orthogonal sequences of characters 8, while the first inputs of the unit for calculating the length of the package of characters 7 and a unit for generating orthogonal sequences of characters 8 are combined, forming the second input of the device, the output of the unit for calculating the length of the packet of characters 7 connected to the second inputs of the symbol packetizer 3 and the orthogonal symbol sequence forming unit 8, the third input of which is connected to the output symbol packetizer 3, the fourth input of the orthogonal symbol sequence forming unit 8 is the third input of the device, the second input of each modulator 5 ₁ -5 _{N is} connected to the corresponding output of the orthogonal symbol sequence forming unit 8.

The unit for the formation of orthogonal sequences of characters 8 (figure 3) is given as an example implementation for the inventive device. The unit for generating orthogonal symbol sequences 8 comprises a driver of N dyadic shifts 9, N parallel branches for forming orthogonal symbol sequences 10 ₁ -10 _N , a storage device 11, a verification unit 12, and a logical element AND 13, wherein the first input of the storage device 11 and the input of the verification unit 12 are combined to form the first input of the block forming the orthogonal sequences of characters 8, the second input of the storage device 11 is the second input of the block forming the orthogonal sequences characters 8, the first input of the driver of N dyadic shifts 9 is the third input of the unit for generating orthogonal sequences of characters 8, each branch of forming orthogonal sequences of characters 10 ₁ -10 _N contains sequentially connected overlay unit of orthogonal codes 14 (respectively 14 ₁ -14 _N in N branches receiving) repeating node 15 (respectively ₁₅ January -15 _n to n receiving branches), the first multiplexer 16 (respectively ₁₆ January -16 _n to n receiving branches) the complex conjugation unit 17 (respectively ₁₇ January -17 _n to n branches n iema) and a second multiplexer (respectively January ₁₈ -18 _N to N receiving branches), the output node overlay orthogonal code 14 (respectively ₁₄ January -14 _N to N receiving branches) is further coupled to second inputs of first 16 and second multiplexers 18, the first inputs of overlay nodes orthogonal codes January ₁₄ -14 _N N branches forming orthogonal symbol sequences coupled with corresponding outputs N shaper shifts dyadic 9, the second inputs of the overlay nodes N orthogonal codes orthogonal branches forming consecutively symbols lnostey January ₁₄ -14 _N are connected to their corresponding outputs of the memory 11, the third inputs of the first multiplexer ₁₆ January -16 _{N N} branches forming orthogonal symbol sequences are combined and connected to the output of the AND gate 13, the third inputs of the second multiplexers January ₁₈ -18 _N N branches of the formation of orthogonal symbol sequences are combined with the first input of the logical element And 13, forming the fourth input of the block forming the orthogonal symbol sequences, the second input of the logical element The tent And 13 is connected to the output of the verification node 12, the outputs of the second multiplexers 18 ₁ -18 _N N branches of the formation of orthogonal sequences of characters are the outputs of the block forming the orthogonal sequences of characters 8.

 The inventive method of diversity transmission of the signal is implemented on the transmitting equipment of a base or mobile station. For this, N diversity channels must be formed on the transmitting side. For example, these channels can be formed by using spatially separated antennas and / or antennas with different polarization. To illustrate the proposed method, an example of its implementation on a BS with N spatially separated antennas is given (Fig. 1).

Packets of M≥N symbols are formed from the stream of transmitted symbols (modulation symbols). An example of this operation is illustrated in FIG. From the symbols of each packet, N orthogonal sequences of symbols s ₁ , s ₂ , ..., s _{N are formed} , each of which contains all the symbols of the packet. Each of the N generated orthogonal symbol sequences of the packet is assigned its own diversity channel, after which a packet is transmitted, in which all the generated orthogonal symbol sequences are transmitted simultaneously. FIG. 5-7 illustrate the above operations of the method, and FIG. 5 illustrates the operations of the method for N = 2, M = 2, FIG. 6 - for N = 4, M = 4 (transmitted characters are real numbers), FIG. 7 - for N = 4, M = 4 (transmitted characters are complex numbers).

 It will be shown below that due to the indicated orthogonality at the receiving side, it is possible to estimate the value of each symbol of the transmitted packet in the absence of the mutual influence of the signals transmitted by different antennas. Moreover, since each symbol of the packet is transmitted in different orthogonal sequences through N different antennas, a diversity effect is present in the symbol estimate.

To form orthogonal sequences from symbols of a packet, the number of symbols in the packet must satisfy the condition M = 2 ^m , where m = 1, 2, .... For example, you can choose M as follows. If N = 2 ^m , m = 1, 2, ..., then M = N. In other cases, you should choose M as the nearest superior number N equal to the power of two.

Под ортогональностью последовательностей s_k,

подразумевается выполнение условия

где К - количество символов в последовательностях s_k;

(•)* - операция комплексного сопряжения.We denote the symbol packet by a = {a _i }, where a _i = А _ci + jA _si is the ith symbol in the packet,

Under the orthogonality of the sequences s _k ,

the condition is met

where K is the number of characters in the sequences s _k ;

(•) * - operation of complex pairing.

 For the formation of N orthogonal sequences, for example, use the following sequence of operations.

операция побитного исключающего ИЛИ. Эту операцию иллюстрирует фиг.5 для N=4, М= 4. Полученные последовательности символов являются линейно независимыми, но не ортогональными. Пример диадических сдвигов исходной последовательности символов в пакете показан на фиг.8.1. N different dyadic shifts of the initial sequence of characters in the packet are performed [12, N. Ahmed, K. R. Rao. Orthogonal transformations in digital signal processing / Per. from English T.E. Krenkel, ed. I. B. Fomenko. M., Communication, 1980]. The result is N character sequences {p _i },

bitwise exclusive OR operation. This operation is illustrated in FIG. 5 for N = 4, M = 4. The resulting symbol sequences are linearly independent, but not orthogonal. An example of dyadic shifts of the original sequence of characters in a packet is shown in FIG.

 The described operation is carried out in the equipment for each packet of characters.

и каждый из коэффициентов u_ij может принимать одно из 2-х значений: +1 или -1. Если символы пакета являются действительными числами, то в результате этой операции формируются N требуемых ортогональных последовательностей символов пакета.2. Place signs in a certain way before sequence symbols. That is, s _i = {p _ij u _ij },

and each of the coefficients u _ij can take one of 2 values: +1 or -1. If the packet symbols are real numbers, then N required orthogonal sequences of packet symbols are generated as a result of this operation.

не зависят от значений символов пакета и могут быть рассчитаны заранее (на этапе проектирования системы связи) для каждого возможного значения длины пакета М. В аппаратуре связи указанные векторы коэффициентов могут быть прописаны, например, в ПЗУ.The coefficient vectors u _i = {u _ij },

do not depend on the values of the symbols of the packet and can be calculated in advance (at the design stage of the communication system) for each possible value of the length of the packet M. In the communication equipment, these coefficient vectors can be written, for example, in ROM.

Рассмотрим теперь несколько примеров нахождения векторов коэффициентов u_i={u_ij},

Первый пример заключается в непосредственном решении системы уравнений (1).Figure 6 shows an example of the formation of N orthogonal sequences of characters from a packet of length M characters, when the characters are real numbers for N = 4 and M = 4. Note that N formed sequences are orthogonal for any values that may have symbols of the packet a _i ,

We now consider several examples of finding the coefficient vectors u _i = {u _ij },

The first example is the direct solution of the system of equations (1).

Коэффициенты u_2j,

последовательности s₂ подбирают таким образом, чтобы обеспечить ортогональность s₁ и s₂. Для этого требуется перебрать всего 2^M значений вектора коэффициентов u₂. Заметим, что может существовать более одного значения вектора коэффициентов u₂, при котором обеспечивается ортогональность указанных последовательностей. Поэтому следует выбрать любое его подходящее значение. На следующем этапе перебирают значения вектора коэффициентов u₃ и выбирают любое такое его значение, при котором обеспечивается ортогональность s₃ с s₁ и s₂. Действуя аналогичным образом, формируют N требуемых ортогональных последовательностей s_i.The second example of finding the values of the coefficients is to sort through the values of the coefficients until a combination of values is found that satisfies the system (1). The maximum number of sorted combinations of the values of the coefficient vectors in this case is 2 ^MN . To reduce computing costs, you can use the following trick. As the first orthogonal sequence s _{1, the} sequence p _{1 is} taken directly, i.e. u _1j = 1,

Coefficients u _2j ,

the sequences s _{2 are} selected so as to ensure orthogonality of s ₁ and s ₂ . To do this, iterate through a total of 2 ^M values of the coefficient vector u ₂ . Note that there can be more than one value of the coefficient vector u ₂ at which the orthogonality of these sequences is ensured. Therefore, you should choose any suitable value. At the next stage, the values of the vector of coefficients u ₃ are sorted out and any such value is selected at which the orthogonality of s ₃ with s ₁ and s _{2 is} ensured. Acting in a similar way, N required orthogonal sequences s _{i are formed} .

Для этого заметим, что искомые векторы коэффициентов u_i являются ортогональными, например, принадлежат к какому-нибудь известному ансамблю ортогональных функций. В качестве последнего можно взять ансамбль функций Уолша [12] {w_i}, w_i={w_ij},

Тогда процедура формирования ортогональных последовательностей s_i выглядит следующим образом. В качестве первого вектора коэффициентов u₁, берут любую из ортогональных функций, например w₁. В качестве u₂ берут такую функцию из оставшихся М-1 ортогональных функций w_i, при которой обеспечивается ортогональность s₁ и s₂. В качестве u₃ выбирают любую ортогональную функцию из оставшихся М-2 функций w_i, такую, чтобы обеспечить ортогональность s₃ с s₁ и s₂. Действуя подобным образом, определяют все N векторов коэффициентов u_i.It is possible to further reduce computational costs by eliminating from the enumeration deliberately inappropriate values of the coefficient vectors u _i ,

For this, we note that the sought vectors of the coefficients u _i are orthogonal, for example, belong to some known ensemble of orthogonal functions. As the latter, we can take the ensemble of Walsh functions [12] {w _i }, w _i = {w _ij },

Then the procedure for the formation of orthogonal sequences s _{i is} as follows. As the first vector of coefficients u ₁ , take any of the orthogonal functions, for example w ₁ . As u ₂ take such a function from the remaining M-1 orthogonal functions w _i , which ensures the orthogonality of s ₁ and s ₂ . As u ₃ choose any orthogonal function from the remaining M-2 functions w _i , such as to ensure the orthogonality of s ₃ with s ₁ and s ₂ . Acting in a similar way, all N vectors of coefficients u _{i are} determined.

3. If the symbols of the packet are complex numbers and N = 2, then replace both symbols in one of the generated sequences with their complex conjugate values or replace one different symbol in each sequence with their complex conjugate values (Fig. 9 positions a - d). The resulting sequences s ₁ and s ₂ are orthogonal.

 If N> 2, then each sequence obtained as a result of the previous operation is repeated, forming N sequences of 2M characters in length. The values of half the symbols in the sequences are changed by the complex conjugate values in such a way as to ensure the orthogonality of all the generated sequences of complex symbols. There are several examples of this replacement.

 1. You can replace the first or last M characters with complex conjugate values in each sequence.

 2. If we consider sequences of characters as rows of a matrix of size Nx2M, then the orthogonality of the rows is ensured by replacing the characters in the M columns of this matrix with complex conjugate values, and the numbers of these columns (counting from zero) taken modulo M should not be repeated. That is, if the first M columns of the matrix are numbered from 0 to M-1 and the last M columns are also from 0 to M-1, then the M columns selected to replace the characters must have different numbers.

 Regarding the formation of orthogonal sequences, the following statements are true.

 1. From the symbols of the packet, M orthogonal sequences can be formed, each of which includes all symbols of the packet. Because N≤M, it is always possible to form N orthogonal sequences required in the method.

 2. The structure of these sequences is such that they are pairwise orthogonal regardless of the meaning of the characters.

 3. When M = 2, the length of the orthogonal sequences is K = 2 characters.

 4. For M> 2, the length of the orthogonal sequences is K = M characters if the symbols of the packet are real numbers, and K = 2M symbols if the symbols of the packet are complex numbers.

 From the above algorithm for the formation of orthogonal sequences of characters, it follows that a large number of their variants can be formed. When implementing the invention, any variant of the formation of orthogonal symbol sequences can be used, but it should be known on the receiving side.

Consider the example of signal reception according to the claimed method of diversity transmission for the case of predominantly single-beam signal propagation in the communication channel, when the dispersion of the time delays of the useful signal at the receiving point is significantly less than the reciprocal of the spectrum of the useful signal. In this case, the propagation channel can be represented as a vector h = [h ₁ (t) h ₂ (t) ... h _N (t)], where h _i (t) is the complex signal transfer coefficient between the ith transmitting and receiving antennas. To receive on the receiving side, it is necessary to know the number N of diversity channels used on the transmitting side, the length M of each symbol packet and the algorithm for generating orthogonal symbol sequences. The receiver may be provided with this information in various ways. For example, all or part of this information may be known in advance, for example, defined by a communication standard. The rest of the information or all of it entirely can be obtained through the service channel in the process of establishing a connection between the base and subscriber stations.

 For reception, estimates of the propagation channels between the transmitting and receiving antennas are also required (estimation of the propagation channel vector h). These estimates are most conveniently generated if a pilot signal is transmitted through each diversity channel. To distinguish the pilot signals at the receiving side, they must be linearly independent. In this case, the best quality of the estimation of the propagation channel can be obtained using orthogonal pilot signals, since in this case, when evaluating the propagation channel between each of the transmitting and receiving antennas, there is no interference from the pilot signals transmitted through the remaining transmitting antennas.

 Two main types of pilot signals are known. This is a continuous pilot signal, which is transmitted continuously during communication simultaneously with the information signal, and an intermittent pilot signal, which is transmitted alternately with the information signal. In the second case, a certain number of pilot symbols known at the receiving side are periodically inserted into the transmitted information symbol stream.

 The pilot signal must be transmitted through the same distribution channel as the information signal. Only in this case, the distribution channel estimate obtained from the pilot signal can be used to receive the information signal. Therefore, the continuous pilot signal must be orthogonal or at least linearly independent with respect to the information signal so that these signals can be separated at the receiving side. This condition can most easily be fulfilled in communication systems with code division of channels, where orthogonal code sequences are used to modulate the signals of various channels. In communication systems with another principle of channel separation, for example with frequency or time division, an intermittent pilot signal is usually used.

 In the case of an intermittent pilot, the transmitted symbol stream contains information and pilot symbols. Moreover, the formation of packets of M symbols is best done in such a way that the symbol packet contains only information or only pilot symbols. In this case, the channel estimation can only be carried out using pilot symbols.

 One of the operations performed at the receiver should be the establishment of packet synchronization, i.e. determination of packet boundaries in the received signal. An algorithm for establishing packet synchronization in a communication system can be implemented in various ways. For example, packet boundaries may be tied to a higher level signal structure. In the IS-95 cellular standard, the signal structure includes the same type of coded character frames of 20 ms duration, synchronized with a period of 1 and Q sequences, as well as with a common system time. Therefore, frame synchronization is easily established in the receiver. In the case of applying the proposed method of diversity transmission, the algorithm for generating each frame can be selected so that each frame includes a known number of packets of M characters. Then frame synchronization can automatically set batch synchronization.

Указанная оценка может быть сформирована по пилот-сигналу любым известным способом. Например, в случае непрерывного пилот-сигнала может быть использован алгоритм оценки комплексной огибающей полезного сигнала из [13, Квазикогерентный прием фазоманипулированного сигнала в системах CDMA/A. В. Гармонов, Ю. Е. Карпитский, И. В. Каюков, В. Б. Манелис//5-я международная научно-техническая конференция "Радиолокация, навигация, связь". Т.1, с. 305-313. Воронеж, 20-23 апреля 1999] . В случае прерывистого пилот-сигнала (пилот-символов) может быть использован, например, интерполяционный алгоритм оценки из [14, S. Sampei and T. Sunaga, "Rayleigh fading compensation for QAM in land mobile radio communications," IEEE Trans. Veh. Technol., vol. 42, pp. 137-146, May 1993]. Будем также считать, что сигналы, пришедшие в точку приема от различных передающих антенн, имеют примерно одинаковые временные задержки.In the reception algorithm considered below, we will assume that an ideal estimate of the propagation channel vector is formed on the receiving side

The specified estimate can be generated by the pilot signal in any known manner. For example, in the case of a continuous pilot signal, an algorithm for estimating the complex envelope of the useful signal from [13, Quasicoherent reception of a phase-shifted signal in CDMA / A systems can be used. V. Garmonov, Yu. E. Karpitsky, I.V. Kayukov, V. B. Manelis // 5th International Scientific and Technical Conference "Radar, Navigation, Communication". T.1, p. 305-313. Voronezh, April 20-23, 1999]. In the case of an intermittent pilot signal (pilot symbols), for example, an interpolation estimation algorithm from [14, S. Sampei and T. Sunaga, "Rayleigh fading compensation for QAM in land mobile radio communications," IEEE Trans. Veh. Technol., Vol. 42, pp. 137-146, May 1993]. We will also assume that the signals arriving at the receiving point from various transmitting antennas have approximately the same time delays.

- полезный сигнал;
s_i,

ортогональные последовательности символов переданного пакета (алгоритм их формирования описан выше); длина каждой последовательности составляет К символов; длительность одного символа равна T_s;

g(t) - некоторая модулирующая функция (в системах CDMA она представляет собой расширяющую спектр сигнала псевдослучайную последовательность); |g(t)| = 1;; n(t) - аддитивный белый гауссовский шум с двухсторонней спектральной плотностью мощностью N₀/2; коэффициенты передачи между передающими и приемной антеннами примерно постоянны на интервале одного пакета h_i(t)≈h_i,

В заявляемом способе разнесенной передачи каждый символ пакета передается на

из К символьных интервалов, которые составляют длительность пакета. При этом на каждом символьном интервале символ передается через соответствующий канал разнесения, с определенным знаком, а на части символьных интервалов передается не само значение символа, а комплексно-сопряженное ему значение. Таким образом, на приемной стороне произвольному m-му

символу принимаемого пакета можно сопоставить последовательность коэффициентов передачи g_m={h_i},

, с которыми он принимается на К символьных интервалах, последовательность знаков e_m={e_mk},

, с которыми символ передавался на указанных интервалах, и вектор ξ_m = {ξ_mk}, ξ_mk = 1, если на k-ом интервале передается прямое значение символа и ξ_mk = -1, если передается комплексно-сопряженное его значение,

Для тех символьных интервалов длительности пакета, на которых m-й символ не передается, соответствующие элементы описанных векторов равны нулю. Для удобства последующей записи введем векторы f_m, f_mk=e_mkg_mk,

На фиг.10 приведены последовательности f_m,

соответствующие примеру реализации изобретения на фиг.7.As a result of standard operations performed in the receiver, an input video-frequency complex signal should be generated [15, Digital radio receiving systems: Reference / M. I. Zhodzishsky, R. B. Mazepa, E. P. Ovsyannikov et al. / Ed. M. I. Zhodzishskogo - M .: Radio and communications, 1990. - 208 p., Ill.], Which in the interval of reception of one packet of symbols [0; T] has the form
x (t) = s (t, a) + n (t), (2)
Where

- useful signal;
s _i

orthogonal sequences of symbols of the transmitted packet (the algorithm for their formation is described above); the length of each sequence is K characters; the duration of one character is T _s ;

g (t) is some modulating function (in CDMA systems, it is a pseudorandom sequence expanding the spectrum of a signal); | g (t) | = 1 ;; n (t) is the additive white Gaussian noise with a two-sided spectral density of power N _0/2 ; transmission coefficients between transmitting and receiving antennas are approximately constant over the interval of one packet h _i (t) ≈h _i ,

In the inventive diversity transmission method, each packet symbol is transmitted to

of K character intervals that make up the duration of the packet. Moreover, at each symbol interval, the symbol is transmitted through the corresponding diversity channel with a certain sign, and not the symbol value itself, but the complex conjugate value is transmitted to the part of the symbol intervals. Thus, on the receiving side, an arbitrary mth

the symbol of the received packet can be associated with a sequence of transmission coefficients g _m = {h _i },

with which it is received at K symbol intervals, the sequence of characters e _m = {e _mk },

with which the symbol was transmitted at the indicated intervals and the vector ξ _m = {ξ _mk }, ξ _mk = 1 if the direct value of the symbol is transmitted at the kth interval and ξ _mk = -1 if its complex conjugate value is transmitted,

For those symbolic intervals of the packet duration at which the mth symbol is not transmitted, the corresponding elements of the described vectors are equal to zero. For the convenience of the subsequent entry, we introduce the vectors f _m , f _mk = e _mk g _mk ,

Figure 10 shows the sequence f _m ,

corresponding to the example implementation of the invention in Fig.7.

Выполним синтез оптимального по критерию максимального правдоподобия алгоритма приема сигнала (2). Логарифм функционала отношения правдоподобия сигнала (2) с точностью до несущественных слагаемых имеет вид

Подставляя (3) и (4) в (5), после необходимых преобразований получаем

где

Δ_k - k-й символьный интервал на интервале пакета,

Положение максимума функционала (6) не изменится, если отбросить несущественный множитель

который на практике всегда отличен от нуля, а также прибавить и отнять от него слагаемое

В результате получаем функционал

Как видно, максимизация функционала (9) эквивалентна независимой максимизации каждого из входящих в него слагаемых, т.е. оптимальным является посимвольный прием. При этом оптимальная оценка m-го символа пакета должна обеспечивать максимум функционала

Поскольку первое слагаемое в (10) не зависит от значения символа а_m, максимизация последнего функционала эквивалентна минимизации второго входящего в него слагаемого. В результате, решающее правило, по которому осуществляется оценка m-го символа, имеет вид

Для относительно простых видов модуляции несложно получить явные оценки информационных символов. Так, для BPSK и QPSK модуляции имеем A_cm=sign(Y_cm); A_sm=sign(Y_sm).Thus, the useful signal (3) can also be represented as follows:

Let us synthesize an optimal signal reception algorithm according to the maximum likelihood criterion (2). The logarithm of the signal likelihood ratio functional (2) up to non-essential terms has the form

Substituting (3) and (4) into (5), after the necessary transformations, we obtain

Where

Δ _k is the kth character interval on the packet interval,

The maximum position of functional (6) does not change if we neglect the insignificant factor

which in practice is always nonzero, and also add and subtract the term from it

As a result, we get the functional

As can be seen, the maximization of functional (9) is equivalent to the independent maximization of each of the terms included in it, i.e. optimal is character-by-character reception. In this case, the optimal estimate of the mth symbol of the packet should provide maximum functionality

Since the first term in (10) does not depend on the value of the symbol a _m , maximizing the last functional is equivalent to minimizing the second term included in it. As a result, the decisive rule by which the mth symbol is evaluated has the form

For relatively simple types of modulation, it is easy to obtain explicit estimates of information symbols. So, for BPSK and QPSK modulation we have A _cm = sign (Y _cm ); A _sm = sign (Y _sm ).

получаем

где

Благодаря взаимной ортогональности последовательностей символов S_i,

справедливы равенства

Раскрывая (12), (13) с учетом (16), (17), получаем
Y_cm+jY_sm=(A_cm+N_cm)+j(A_sm+N_sm); (18)
При фиксированном векторе канала h шумовые слагаемые N_cm и N_sm являются некоррелированными гауссовскими случайными величинами с нулевыми средними и с одинаковой дисперсией, равной

Следовательно, ОСШ в решающей статистике m-го символа пакета определяется выражением

Как видно, выражение для ОСШ на символ при использовании заявляемого способа разнесенной передачи имеет такую же структуру, как и при оптимальном когерентном разнесенном приеме сигнала по N каналам разнесения [1]. Это говорит о высокой эффективности изобретения.Let us analyze the value of the signal-to-noise ratio (SNR) in the decisive statistics Y _cm + jY _{sm of the} mth symbol. To do this, expand the expressions for Y _cm and Y _sm . Substituting (3) into (7) and considering that

we get

Where

Due to the mutual orthogonality of the sequences of characters S _i ,

fair equalities

Revealing (12), (13), taking into account (16), (17), we obtain
Y _cm + jY _sm = (A _cm + N _cm ) + j (A _sm + N _sm ); (18)
For a fixed channel vector h, the noise terms N _cm and N _sm are uncorrelated Gaussian random variables with zero averages and the same dispersion equal to

Therefore, the SNR in the decisive statistics of the mth packet symbol is determined by the expression

As you can see, the expression for SNR per symbol when using the proposed method of diversity transmission has the same structure as in the case of optimal coherent diversity reception of a signal over N diversity channels [1]. This indicates the high efficiency of the invention.

 When K = M, the implementation of the proposed method does not increase the channel transmission rate (the repetition rate of characters transmitted through the communication channel), compared with the repetition rate of characters in the input stream (at the input of the proposed method). This is an important advantage of the proposed method in comparison with the known CTD method. In the case when N> 2 and the transmitted characters are complex numbers, K = 2M, and thus the implementation of the proposed method leads to a twofold increase in the frequency of the characters at the output of the proposed method compared to the frequency in the input stream of characters. However, in this case, the inventive method has a great advantage compared to CTD, similar in noise immunity, since the implementation of the latter leads to an N-fold increase in the symbol repetition rate.

 The inventive method is implemented on the device, a block diagram of which is shown in figure 2, 3.

 The first inputs of the unit for calculating the length of the packet of symbols 7 and the unit for generating orthogonal sequences of symbols 8 receive a signal on the number of transmitting antennas at the base station. The output signal from block 7 (the length of the packet of characters) is supplied to the second inputs of block 3 and block 8.

The input symbol stream goes to the first input of the symbol packet generator 3. In accordance with the input signals, block 3 generates symbol packets that go to the third input of the block for generating orthogonal symbol sequences 8. A signal of the modulation type (real or quadrature) is received at the fourth input of block 8. Block 8 generates N orthogonal symbol sequences for each symbol packet. N orthogonal symbol sequences are supplied from the output of block 8 to the inputs of their respective modulators 5 ₁ -5 _N (the first orthogonal symbol sequence is fed to the input of the first modulator, the second to the input of the second modulator, etc.). The second inputs of the modulators 5 ₁ -5 _N receive output signals from the block forming the pilot signals. In modulators, the operations of modulating the carrier oscillation by an input signal, amplification, filtering, etc. The output signals of the modulators are transmitted to the transmit transmit antennas 6 ₁ -6 _N.

 Consider the formation of orthogonal sequences of characters in block 8 (figure 3).

 The signal about the number of transmitting antennas is fed to the first input of the storage device 11, as well as to the input of the verification unit 12. The signal of the symbol packet length is received at the second input of the block 11.

The symbol packet is fed to the input of the shaper of N dyadic shifts 9. Block 9 performs N different dyadic symbol shifts in the packet, thus forming N sequences of symbols that arrive at the first inputs of N nodes of the overlapping orthogonal code 14 ₁ -14 _N (the first sequence of characters goes to a first input node of the first orthogonal code overlay 14 _1, the second sequence - the first input of the second node _N 14, etc.) the second inputs of N orthogonal code overlay nodes on January ₁₄ -14 _N receives the output signals from the memorizing on the device 11 (orthogonal codes). At nodes 14 ₁ -14 _N , orthogonal codes are superimposed on the generated symbol sequences. The sequences of symbols formed at the nodes 14 ₁ -14 _N from the outputs of these nodes go to the first inputs of the corresponding repetition nodes 15 ₁ -15 _N , as well as to the second inputs of the multiplexers 16 ₁ -16 _N and 18 ₁ -18 _N. In the repetition node, each input sequence of characters of length M characters is repeated, the result is a sequence of 2M characters that is output from this node. Note that the repetition operation implemented in repetition nodes is performed only for complex characters. The third inputs of the multiplexers 16 ₁ -16 _N receives the output binary signal from the logical element And 13. Depending on the value of this signal, the outputs of the multiplexers 16 ₁ -16 _N receive sequences of characters from the first or second inputs. The sequence of characters from the outputs of the multiplexers 16 ₁ -16 _N goes to the inputs of the nodes of the complex interface 17 ₁ -17 _N. In the nodes of complex conjugation 17 ₁ -17 _N change the values of some of the characters in the input sequences to complex conjugate values in such a way as to ensure the orthogonality of all generated sequences of characters regardless of the values of the characters of the package. Orthogonal sequences of characters from the outputs of the nodes of complex conjugation 17 ₁ -17 _N are supplied to the first inputs of the multiplexers 18 ₁ -18 _N , the third inputs of which receive a signal of the form of modulation. Depending on the value of this signal, the outputs of these multiplexers and, accordingly, the outputs of block 8 receive orthogonal sequences of symbols from the first inputs (complex characters) or from the second inputs (real characters) of the multiplexers 18 ₁ -18 _N.

 A comparative analysis of the characteristics of the claimed invention and known technical solutions in the art was carried out. Comparison of different diversity signal transmission methods was performed for the CDMA IS-95 direct channel signal format [15, Subscriber and base station compatibility standard for dual-mode cellular broadband systems with TIA / TIA / IS-95-A spread spectrum, may 1995. Telecommunications Industry Association ]. The results are shown in Figs. 11-13 in the form of dependences of the probability of erroneous reception of the IS-95 direct channel frame on the SNR per bit in the receiver of the subscriber station. The frame size was equal to 192 bits, of which: 172 information, 12 CRC bits and 8 tail, which corresponds to a transmission rate of 9.6 kb / s. Each frame was encoded with a convolutional code with the following parameters: coding rate 1/2, code restriction 9, encoder polynomials 0753, 0561. Modeling of PSTD, OTD and according to the invention was performed at a symbol level, i.e. the operations of superimposing the transmission of the signature signal code of the user and the expanding signal spectrum of the BS pseudorandom sequence were not essential for testing the indicated methods of diversity transmission. When modeling DTD, these operations are significant, and therefore, the BS signal was modeled exactly in accordance with the IS-95 standard. At the same time, the signals of 32 users in a cell were simulated. The simulation was performed for the conditions of independent flat Rayleigh fading of signals arriving at the receiving point from different transmitting antennas. The fade rate was 30 Hz. Fading generators were designed according to Jakes model [16, Casas and Leung, “A Simple Digital Fading Simulator for Mobile Radio,” 38th IEEE VTC, June 1988, pp. 212-217]. The interference was simulated by a Gaussian band noise. The computer model of the receiver included one receiving antenna. Estimates of the temporal position and complex envelopes of signals coming from different antennas were assumed to be perfect. The decoding of the received frames was carried out according to soft decisions in accordance with the Viterbi algorithm.

 The PSTD, DTD, OTD and “invention” curves of FIGS. 11-13 are obtained for respective diversity transmission methods. The curves "stationary channel" and "one transmit antenna" are obtained for one transmit antenna, the first in the absence of fading, and the second in the presence of fading in the communication channel.

 As follows from a comparison of the curves of FIGS. 11–13, the claimed diversity transmission method has the best noise immunity. At N = 8, the inventive method provides almost the same noise immunity of reception in the forward channel as in the absence of fading in the communication channel. In the work area (the probability of a frame error of about 1%), the energy gain of the claimed invention compared to the transmission algorithm through a single antenna is 4, 7 and 9 dB for 2, 4 and 8 transmitting antennas, respectively. Considering that the capacity of CDMA systems is inversely proportional to the SNR value at which the required quality of communication is ensured, the corresponding increase in the capacity of the forward channel is 2.5, 5, and 8 times.

 High noise immunity of the proposed method of diversity transmission of a signal and a device for its implementation are achieved due to the efficient diversity at the level of each transmitted symbol and the complete elimination of mutual interference between the symbols of the received packet, which is achieved due to the orthogonality of the sequences of symbols transmitted over the diversity channels. The implementation of the invention on the transmitting and receiving side does not require significant costs.

Claims (9)

 1. A method of diversity transmission of a signal, which consists in generating N diversity channels of signal transmission, packets of N symbols are formed from the input stream, characterized in that packets of M> N symbols are formed, if N is not equal to the power of two, from symbols each packet form N orthogonal symbol sequences in such a way that each of them contains all the symbols of the packet, and the rule of forming orthogonal symbol sequences should be known at the receiving side, assign the diversity channel each generated orthogonal symbol sequence of each packet and carry out sequential transmission of the packet.  2. The method according to claim 1, characterized in that the input symbol stream contains information and pilot symbols.  3. The method according to claim 1, characterized in that packets of M symbols are formed from the input symbol stream so that each packet consists only of information symbols or of pilot symbols.  4. The method according to claim 1, characterized in that when forming packets of M symbols, M ≥ N is selected if the number of diversity channels is equal to a power of two; in the opposite case, M is chosen to be the nearest number N that is equal to a power of two.  5. The method according to claim 1, characterized in that, forming N orthogonal symbol sequences, N different dyadic shifts of the original symbol sequence of the packet are performed, forming N symbol sequences.  6. The method according to claim 1 or 5, characterized in that if the symbols of the packet are real numbers, then, forming N orthogonal symbol sequences, multiply the symbols in the sequences formed by the coefficients + A or -A, where A is a non-zero real a number, in such a way as to ensure the orthogonality of all generated symbol sequences, regardless of the symbol values of the packet.  7. The method according to claim 1 or 5, characterized in that if the symbols of the packet are complex numbers and N = 2, then, forming orthogonal sequences of characters, the characters in the resulting sequences of characters are multiplied by coefficients + A or -A, where A is a non-zero real number, and change the values of part of the characters to complex conjugate values in such a way as to ensure the orthogonality of the generated sequences of characters regardless of the values of the characters of the package.  8. The method according to claim 1 or 5, characterized in that if the symbols of the packet are complex numbers and N> 2, then, forming N orthogonal symbol sequences, each formed sequence of symbols is repeated, forming N symbol sequences of 2M symbols long, the symbols are multiplied in the formed sequences of characters by the coefficients + A or -A, where A is a non-zero real number, and change the values of some of the characters by the complex conjugate values in such a way as to ensure orthogonally All sequences be formed independently of the values of packet symbols.  9. A diversity signal transmission device comprising a symbol packetizer, a pilot generator, N modulators and N transmit antennas, the first input of the symbol packetizer being the first input of the device, the first input of each modulator connected to the corresponding output of the pilot signals, the output of each modulator is connected to its corresponding antenna, characterized in that a block for calculating the length of the symbol packet and a block for generating orthogonal sequences of symbols are introduced tins, while the first inputs of the block for calculating the length of the symbol packet and the block for generating orthogonal symbol sequences are combined to form a second input of the device, the output of the block for calculating the length of the symbol packet is connected to the second inputs of the symbol packetizer and the block for generating orthogonal symbol sequences, the third input of which is connected to the output symbol packer, the fourth input of the orthogonal sequence of characters is the third input of the device, the second input q of each modulator is connected to the corresponding output of the block for generating orthogonal symbol sequences.

Images