GB2372411A

GB2372411A - A process for multi-sensor equalisation in the presence of interference and multiple propagation paths

Info

Publication number: GB2372411A
Application number: GB9504244A
Authority: GB
Inventors: Francois Pipon; Pascal Chevalier; Pierra Vila
Original assignee: Thomson CSF SA
Current assignee: Thales SA
Priority date: 1994-03-04
Filing date: 1995-03-02
Publication date: 2002-08-21
Anticipated expiration: 2015-03-02
Also published as: DE19507473C2; ITTO950153A1; FR2735635B1; GB9504244D0; IT1279152B1; GB2372411B; FR2735635A1; DE19507473A1; ITTO950153A0

Abstract

Multi-path signals received at a number of sensors C1-CN are transformed to baseband, sampled at a multiple of the symbol rate and low pass filtered. The signals are combined 14 to produce an adaptive array directing the main gain towards the main source and then re-sampled at the symbol rate by T<SB>S</SB>. Spatial filtering W<SB>1</SB> W<SB>2</SB> separated by delays equivalent to those of selected multi-paths, extract useful signals and input them to calculator 16 which provides the inputs to a transverse filter part of temporal filter T. The re-sampled signal Zs(n) is added to the transverse filtered signal Z<SB>T</SB>(n) and input to recursive stage R. The filter coefficients are continuously calculated by an adaptive algorithm seeking to minimise the error in the receiver output d(n).

Description

237241 1

A PROCESS FOR MULTI-SENSOR EQUALISATION IN A

RADIO RECEIVER IN THE PRESENCE OF

INTERFERENCE AND MULTIPLE PROPAGATION PATHS

The invention relax to a praxes for multi-sensor equalization in a radio receiver consisting of the demodulation of a digital message in the presence of multiple propagation paths and sources of interference by reducing the number of coefficients to be adapted required for the calculation lo of a multi-sensor equaliser for modulations composed of learning sequences and information symbol sequences. The invention also concerns a radio receiver implementing such a process.

It applies to the high frequency (HF) band, which is particularly important in the field of radio communications in that it enables long distance

> communication based on phenomena of reflection off the different layers of the ionosphere. Such communication is based on antenna processing techniques, and therefore requires the use of an array consisting of several sensors. In many digital radio communication applications transmission between go the transmitter and the receiver is carried out along several propagation paths. As the delay time between the different paths can be greater than the symbol duration, equalization becomes necessary to compensate for inter-

symbol interference (ISI) which can be generated.

This phenomenon is particularly likely to occur in the HF band, where at, the multiple propagation paths produced by reflection off different layers of the ionosphere can be spaced by up to 5 ms, i.e. several times the symbol duration in the case of modulations whose bandwidth is typically of the order of 3 kHz. It also occurs in other frequency bands in communications using very high bit rates, such as GSM (270 kbitis, i.e. a symbol duration of 3.7 As), 30 in urban or mountainous terrain, where the path differences produced by reflection off obstacles (buildings, mountains, etc.) can be separated by up to 10 or even 20 as.

Many systems currently in use are capable of adapting to these propagation conditions by the insertion in the wave form of learning

sequences which are known to the receiver. This leaves a range of possible solutions for the adaptive equalization of the useful signal received.

A first solution consists in using a Viterbi algorithm which requires a prior estimation of the propagation channel using the learning sequence. This s equalization method has the advantage of minimizing the probability of error across the whole sequence of information symbols, but becomes very costly when the duration of the pulse response of the channel is much greater than the symbol duration. In fact, the number of states that the Viterbi algorithm must process is equal to ML, where M is the size of the modulation alphabet to and L is the length of the pulse response of the channel expressed as a number of symbol periods. This solution is used for GSM type applications where the Viterbi algorithm typically consists of 32 states (L = 5 and M = 2).

In the HE band, the particular field of application of the invention, the

number of states becomes too great for the Viterbi algorithm to be practically realisable (typically, M is 4 or 8, while L is equal to 12, which corresponds to a pulse response spread over 5 ms), and a second solution using a DFE equaliser is often used.

This second solution consists in using learning sequences as the response of an adaptive algorithm used to minimise a MQE (Mean Quadratic JO Error) criterion. This solution uses a "Decision Feedback Equaliser" (DFE).

Such an equaliser is intended to supply to a decision module adapted to the modulation in question a signal in which ISI has been eliminated or at least reduced to a great extent. To this end, the DFE equaliser uses transverse and recursive self-adapting filters which are adapted by a least hi, squares type algorithm, preferred to a gradient algorithm for reasons of speed of convergence. The known symbols in the learning sequences are used for the adaptation of the different coefficients. The tracking of channel variations beyond the known sequences is effected using symbols which are selected (detemined) as responses as necessary during the execution of the process.

JO The single-sensor DFE equaliser can compensate for ISI caused by multiple propagation paths, but is not capable of phase realigning these different paths. Thus, in the presence of two stationary paths of the same amplitude, the DFE equaliser produces losses of approximately 3 dB with respect to a white Gaussian noise channel: it endeavours to retain the

contribution of one of the paths and to eliminate the second using the recursive part.

Moreover, in the HE band the different propagation paths are very often affected by flat fading. Fading is a phenomenon linked to the variation of 5 the multiple paths which in turn produces a variation of the received power, or even in extreme cases fading or dying out of the signal paths. When fading is strong, a DEE equaliser's performance is seriously reduced.

In addition, these techniques rapidly become inefficient in the presence of jamming, which means that it is necessary to use known specific lo anti-jamming techniques such as error correction encoding, elimination of jamming by notched filtering, use of frequency evasion links, etc.. These techniques are used in many operational systems, but are nonetheless of limited effect when interference is strong and occupies the whole of the useful signal band. In such conditions, it is necessary to use more effective anti-jamming means based on the use of antenna filtering techniques.

Antenna filtering techniques appeared in the early 1 960's. One in particular is described in an article of P.W. HOWELLS "Explorations in fixed and adaptive resolution at GE and SURC", IEEE Trans-Ant-Prop, vol. AP-24, no. 5, pp 575-584, Sept. 1976, while an exhaustive synthesis is presented in so a doctorate thesis presented by P. CHEVALIER at the University of Paris sud in June 1991 entitled "Antenne adaptative: d'une structure lineaire a one structure non lineaire de Volterra" ("The adaptive antenna: from a linear structure to a non-linear Volterra structure"). These techniques are designed to combine the signals received by the various sensors making up the 95 antenna so as to optimise its response to the useful signal and jamming scenario in question.

The selection of sensors and their disposition is an important parameter which has a central influence on the performance of the system.

Three basic configurations are possible: so - the sensors are identical and disposed at different points in space, discrimination between the useful signal and interference being effected according to the direction of arrival; - the sensors are all disposed at the same point in space (colocalised antenna) and have different radiation diagrams. This means that

discrimination can be carried out on the basis of polarization and direction of arrival; -the two configurations described above can be combined: several colocalised antennas can be disposed at different points in space.

5 In addition, since propagation and jamming conditions can change over time, it is essential that the system be capable of adapting the antenna to these variations in real time through the use of a particular antenna filtering technique: the adaptive antenna. An adaptive antenna is one which detects and reacts to sources of interference automatically by constructing holes in its to radiating diagram in their direction, while at the same time improving reception of the useful source, without any prior knowledge of the interference and on the basis of a minimal amount of information on the useful signal.

Moreover, the tracking capabilities of the algorithms used make an adaptive antenna able to respond automatically to a changing environment.

Adaptive antennas are characterized by the way in which they discriminate between the useful signal and interference, i.e. by the nature of the information relating to the useful signal which they use. This discrimination process can be carried out in one of five different ways: according to direction of arrival, TO - according to modulation, according to time, for example, with frequency evasion links, - according to power, - blindly (for example, higher order source separation methods).

Up until very recently, transmission systems have always been based ?5 upon the independent operation of single-sensor adaptive equalization and adaptive antenna techniques, which results in less than optimised performance. Thus, the system described in an article by R. Dobson entitled "Adaptive antenna array", patent no. PCT/AU85/00157 of February 1986, so which uses discrimination according to time, is efficient in terms of interference rejection, but makes no attempt to improve the useful signal to noise ratio.

In a transmission context, and when learning sequences are introduced into the wave form, it is preferable to use antenna processing techniques

based on discrimination according to modulation, as these techniques enable optimization of the useful signal to noise ratio. Most techniques used nowadays attribute complex weightings to each of the sensors of the adaptive antenna. Such an antenna is capable of rejecting interference, but in the 5 presence of multiple propagation paths: - it "aligns" on the direction of one of the paths, i.e. it phase realigns the contributions of this path on the various sensors (for omnidirectional sensors, a signal to noise ratio gain of 10 log N is obtained, where N is the number of sensors used), lo - it also attempts to eliminate non-correlated paths from the signal, thus losing the energy associated with these paths.

In order to improve the performance of this type of antenna processing in the presence of multiple propagation paths, it is possible to combine it with a single-sensor equalization technique to obtain a multi-sensor equaliser 15 consisting of a spatial part, composed of different filters disposed on each of the reception channels, and a temporal part located at the output of the spatial part. All the filters making up the spatial part and the temporal part are jointly adapted to the same error signal.

Several multi-sensor equalisers have already been proposed and so studied, principally in the field of mobile radio transmissions, and these are

particularly described in an article by K.E. Scott and S.T. Nichols entitled: "Antenna diversity with Multichannel Adaptive Equalization in Digital Radio" and in an article by P. Balaban and J. Salz entitled "Optimum Diversity Combining and Equalisation in Digital Data Transmission with Applications to 95 Cellular Mobile Radio - Part 1: Theoretical Considerations", IEEE Trans. on Com., vol. 40, no. 5, pp 885- 894, May 1992.

Up to now, such equalisers have been intended to combat the selective fading engendered by multiple paths in a non-jammed environment.

They consist of Finite Pulse Response filters, one on each channel, followed so by an adder then a monodimensional equaliser equalising at the symbol rate.

The criterion used for the optimization of these multi-sensor equalisers is the minimization of MOE between their output and a response determined by the learning sequences.

In the equaliser proposed by Scott et al, coefficient adaptation is carried out by a least squares algorithm, and its use for a HE channel cannot be envisaged given the wave forms used. Taking into account the temporal spread of the multiple paths, the number of coefficients to be adapted is too great for the algorithm to be able to converge with the learning sequence.

The aim of the invention is to resolve these problems.

Accordir to one aspect of t)liS irn ention there is provided a process for mllti sereor equalization in a rnd:io receiver c isiry: a static part connected to a temporal part composed respectively of a determined number of filters and lo receiving a radio signal consisting of at least a learning sequence made up of symbols known to the receiver and an information sequence made up of useful symbols, and caq risir g in a first stage for the preliminary processing of the signal received by the receiver, of the transformation of the signal received by at least two sensors into an equivalent baseband signal, of the 15 sampling of the baseband signal at a rate which is a multiple of the symbol rate and of the filtering of the sampled signal using a low-pass filtering process, wherein, in the presence of interference and multiple propagation paths, it consists, in order to reduce the number of filter coefficients to be adapted, in a second stage, of a synchronization measurement process, of go the estimation of the number of paths in the signal, the delay times associated with the various paths and their relative powers, and the frequency offset between the emission and reception of the signal in order to compensate for it, and wherein, in a third stage of multi-sensor equalization, it consists in selecting a determined number of paths according to a determined at, criterion from the number of paths estimated in the synchronization stage, in filtering via a spatial processing procedure the signal received by the receiver using the filters of the spatial! part, in filtering via a temporal processing procedure the signal output by the spatial part using the filters of the temporal part, the respective coefficients of the filters of the spatial part and the to temporal part being jointly and periodically recalculated at each iteration by an adaptive algorithm working at the symbo! rate in order to minimise the estimation error produced between the receiver output signal and the response signa!

According to another aspect of this invention there is pewided a radio receiver featuring at least one multi-sensor spatial diversity eq'=li.q r consisting of a spatial part connected to a ten penal part and receiving a digital radio signal composed of at least a learning sequence made up of symbols known to the receiver and an 5 information sequence made up of useful symbols, featuring, in order to reduce the number of filter coefficients to be adapted in the spatial and temporal part in the presence of interference and multiple propagation paths: - at least two sensors connected to a unit carrying out the preliminary processing and synchronization of the receiver input signal, the outputs of this lo unit being connected respectively to a first series of inputs and a second series of inputs of the spatial part of the equaliser, the first series of inputs corresponding respectively to the inputs of the spatial filters relating to each of the paths selected from a determined number of paths detected, and the second series of inputs corresponding respectively to the inputs of a unit for is the calculation of the input signals of the transverse part of the temporal part of the equaliser, and wherein the transverse part of the temporal part features a transverse filter of determined coefficients, the temporal part also featuring a recursive part consisting of a decision module whose output is connected to the input of a recursive filter of determined coefficients, the recursive filter go being located in a loop and receiving on its input the sum of the output signals of the spatial part and of the transverse part, from which is subtracted the signal output by the recursive part.

The process in Which the present invention is embodied on the one hand enables an improvement upon the performance of the various single-sensor equalizers as currently in existence: in the case of a stationary environment, the process according to the invention enables an improvement of 10 log N in antenna gain, where N is the number of sensors, where the sensors are identical, with a gain of 3 dB on the phase realignment of the paths in the case of two stationary paths of the same power.

so Moreover, the multi-sensor equalization process in which the invention is embodied improves to an even greater extent the performance of singlesensor equalization in the presence of flat fading on the various propagation paths.

The structure of a receiver in which the invention is embodied using a single-sensor equaliser also greatly reduces the number of coefficients to be adapted in comparison with the structure proposed by Scott et al, and can therefore be implemented on a HE or GSM channel.

s Other characteristics and advantages of the invention will be made clear in the following description, accompanied by the appended figures

which represent, respectively: - figure 1, the main stabs of the process in which the invention is Bodied; - figure 2, the main stages of the preliminary processing stage; lo - figure 3, the main stages of the synchronization stage of the process in which the invention is embodied; figure 4, the main stages of the multi-sensor adaptive equalization stage of the process in which the irTv ntion is embodied; - figure 5, an example structure of a radio receiver in which the -> invention is embodied; figure 6, an algorithm of the spatial matrix used by the process in which the invention is embodied; - figure 7, an array of antennas used by the receiver in whirs the invention is embodied; ?0 - figure 8, a graphic representation demonstrating the importance of the spatial part in the structure of the equaliser of the receiver in which the invention is embodied.

Stage 1 of the prisons in which the invention is embodied represented in figure 1 consists of the preliminary processing of a digital signal received by at least hi, two sensors On, where n = 1 to N. of a radio receiver.

Stab 2 of the process in which the invention is embodied consists in synchronizing the pre-processed received signal with an emitted signal consisting of synchronization sequences known to the receiver in the presence of interference and multiple paths.

so Synchronisation stage 2 necessarily precedes multi-sensor equalization stage 3, which consists of spatial processing of the signal followed by temporal processing, both processing procedures being jointly adapted.

Preliminary processing stage 1 is subdivided into three main stages 4, 5 and 6 as illustrated in figure 2: - stage 4 consists of the transformation of the radio signal output by sensors On into a baseband signal; - stage 5 consists in sampling the transformed baseband signal at a rate Te, Te being a multiple of the symbol rate Ts, and - stage 6 consists in filtering the sampled signal using a low-pass filtering procedure. The pre-processed and synchronous signal derived from stages 1 and lo 2 is subsequently referred to as " the signal output by the reception channels ".

Multi-sensor synchronization stage 2 is subdivided into three main stages 7, 8 and 9 as illustrated in figure 3: - stage 7 consists in measuring the synchronization of the signal received by 15 the senators against learning sequences made up of symbols known to the receiver; - stage 8 consists in estimating of the number of paths followed by the useful signal as well as the delay times associated with these paths and their relative powers, and no - stage 9 consists in estimating the frequency offset between emission and reception. This frequency offset is compensated for before the multi-sensor equalization stage is carried out.

Multi-sensor adaptive equalization stage 3 is subdivided into five main stages 10 to 12 as illustrated in figure 4.

25 In stage 10 the process in which the inv ticn is mbodied *Moses to adapt to K paths selected from the P paths identified at the end of synchronization stage 2. This selection can be based on a number of different criteria: - limit the number of coefficients of the spatial processing component of the equalization for reasons of calculation power or optimization of the So convergence speed, by imposing, for example, K 2; - select all paths of which the relative power with respect to the main power is sufficiently great for phase realignment to be beneficial, for example a relative power of -5 dB; - use the two criteria described above simultaneously and concurrently.

1 0 Spatial proc ssing stage 11 consists in filtering the input signal using filters disposed on each of the sensors making up the array, and phase realigns the contributions of all the paths selected, provided that these are sufficiently spaced in spatial terms, which implies a coefficient of spatial 5 correlation between the different directing vectors which is " sufficiently " less than 1, as well as the positioning of the gain of the antenna in the direction of the useful signal.

Spatial processing stage 11 also rejects interference.

Temporal processing stage 12 consists in filtering the signal output by to spatial processing stage 11 with a filter consisting of a transverse part and a recursive part, and combats any ISI remaining after the spatial processing of stage 11 due either to paths not selected in the algorithm or to paths which are spatially too close to one another for the spatial processing of stage 11 to be capable of separating them.

5 The coefficients of the filters used for the spatial and temporal processing associated respectively with stages 11 and 12 are jointly adapted to the symbol rate Ts by the adaptation algorithm so as to minimise MOE between a response signal and the result of equalization stage 3. The response signal consists either of known symbols belonging to a learning so sequence or of " determined " symbols where the symbol in question belongs to an information sequence.

A radio reopener in Rich the Rich is "bodied r ceivir a digital signal including learning sequences and information sequences is schematically illustrated in figure 5.

95 1 receiver implments the process in Rich the i entia is Lied and the description which follows is intended to aid understanding of its operation.

An emitted signal d(t) arrives at a reception array of a receiver in Rich the irrventi is Readied feal rir g a deterred robber of servitors en, where n = 1 to N. after its journey through the ionospheric channel. Each of So the P propagation paths followed by the signal is received by the antenna with a complex gain X(t) and undergoes a delay hi with respect to the emitted signal. The vector X(t) formed by the signals received by the sensors is determined by the following formula:

1 1 pa X(t)= >,aj(t)d(t-Ij)Sj+B(t) (1) i=1 where: Sj represents the direction vector associated with path i, andB(t) is added noise independent of the useful signal which takes into account the contributions of background noise and interference.

5 The non-stationary nature of the channel affects the amplitudes and phases of the various paths, hence the dependence in time of the quantities a;(t). On the other hand, the delays Ij are relatively stable over periods of the order of a quarter of an hour and can therefore be considered constant.

Sensors Cn are respectively connected to the input of a preliminary lo processing and synchronization unit 13 featuring conventional means, which are not represented, for the transformation of the signal received by sensors Cn into a baseband signal, its sampling at the rate Te, its transformation into a baseband signal and its low-pass filtering, as well as conventional means for synchronization in the presence of jamming. Each output of unit 13 15 corresponds to the reception channel associated with one of sensors Cn, and each supports a part of the complex baseband signal sampled at rate Te.

The estimated delay times can be expressed as a function of Te: = pi Te, and the sampled signal X(nTe) received by the antenna can thus be written as follows: p no X(nTe) = 2, ajd(nTe - pjTe)Sj + B(nTe) (2).

i=1 The structure of the multi-sensor equaliser connected at its output to unit 13 consists of a first part termed the " spatial part " and a second part termed the "temporal part". The dimensions S of the spatial part, which defines the number of coefficients required for its calculation, are determined 25 by the product of the number K of paths selected at the end of stage 6 and the number N of sensors C1 to ON. The spatial part rejects any interference and positions the gain of the antenna equivalent to the array of sensors Cn in the direction of the useful signal, and if possible realigns the phase of the multiple paths associated with the useful signal.

so in a conventional multi-sensor equaliser such as that proposed by Scott et al, the spatial part consists of a Finite Pulse Response filter, or FPR,

l 1 2 disposed on each reception channel. Each filter consists of a determined number of coefficients so as to cover the whole of the transmission channel.

Each of these coefficients is represented in figure 5 by a broken line box. To cover a channel whose length in the HE band can be typically 5 ms, and with 5 sampling at 3 kHz, the number of coefficients required on each channel is 3 x 5 = 1 5.

1h the sots part of the receiver in Will Lee inv ticm is erbodied, the number of coefficients to be adapted is greatly reduced. Only K coefficients per channel, typically one, two or even three in the HF band, need to be lo calculated. Each coefficient selected is represented in figure 5 by a solid line box (K= 2 in figure 5). These K coefficients per channel enable the definition of K vectors, each of these vectors respectively forming a vertical spatial filter Wk. where k = 1 to K, represented by a solid line. Each of these filters Wk weights a signal vector Xk(n).

15 Xk(n) is defined as the vector which enables the symbol d(n) on path k to be taken into account by the equaliser at moment n.

This structure therefore reduces the number of coefficients of the spatial part. The outputs of filters Wk are summed by a first summing circuit 14 whose output, which delivers the signal Zs(n), is connected to a first so positive operand input of a first comparator 15, which also corresponds to a first input of the temporal part.

It should be noted that synchronization step 2 has been carried out by oversampling the input signal d(t) with respect to the symbol rate, which enables the delays of the various paths to be estimated with greater precision 95 at the synchronization stage, and therefore means that the maximum possible amount of energy is recovered on each of the paths selected subsequently at multi-sensor equalization stage 3.

The precision of the estimation of the delays is therefore particularly important in optimising the performance of the multi-sensor equaliser of the so receiver according to the invention. In addition, the structure is not fixed, and synchronization step 2 means that the spatial part of the structure can be monitored and updated when one of the paths disappears (fading hole) or appears, or when the delay times are modified, for example in the case of clock drift between emission and reception.

1 3 The spatial part also features a unit 16 for the calculation of the input signals of a first part of the temporal part termed the "transverse part".

Calculation unit 16 receives the signals output respectively by the preliminary processing and synchronization unit 13 on a first series of inputs, and s receives signal vectors Xk(n) output respectively by filters Wk on a second series of inputs.

The operation of unit 16 is described in detail below.

The transverse part is designed to compensate for inter-symbol interference (ISI) remaining in the signal at the output of the spatial part.

to The transverse part receives the signals delivered by calculation unit 16 and features a transverse filter with T coefficients hereafter termed "HT".

The outputs of filter HT are summed by a second summing circuit 17, whose output, which delivers the signal zT(n), is connected to a first positive operand input of comparator 15.

15 The output of comparator 15 is connected to a first input of a second part of the temporal part termed the "recursive part". The recursive part consists of a decision module 18 situated in a main circuit and a recursive filter hereafter termed "HR", with R coefficients, situated in aloop. This filter HR receives the signal delivered by the decision module on its input, and its so output signal is delivered to a third negative operand input of comparator 15.

The output of comparator 15 is on the one hand returned to the input of decision module 18 and on the other hand delivered to a first positive operand input of a second comparator 20, which receives on a second negative operand input the response signal also termed response d(n). The output of 5 second comparator 20 delivers a minimised error estimation signal e(n).

The output of the temporal part delivers the " determined " symbols.

The spatial and temporal parts are jointly adapted to the symbol rate Ts represented by a switch located between summing circuit 14 of the spatial part and the temporal part in such a way as to minimise a MQE criterion so between the response signal also termed response d(t) and the multisensor equaliser output signal z(t).

Ideally, the optimised criterion for the calculation of the various filters Wk. HT and HR making up the structure is a criterion of MQE between output signal z(t) and response d(t). It is determined by the following formula:

= E[lz(t)- d(t)l] (3).

Given that the statistics relating to the signals are not precisely known, the calculation of different filters Wk. HT and HR is carried out using an adaptive algorithm operating at symbol rate Ts and optimising for each 5 iteration, i.e. for each sample n, a MQE criterion estimated using the following formula: 1 n t(n) =-2,|z(i) - d(i)|2 (4).

i=1 The adaptive algorithm is here defined for a stationary channel, and converges on the solution which obtains the minimum MQE between d(t) and lo z(t). In a non-stationary environment, the algorithm minimises MQE over a short period related to the degree to which the channel is nonstationary. This is achieved by weighting the MQE samples with a window which is generally exponential. The criterion to be minimised for each sample is determined by the following formula: i:;.(n) = 2,tn i|z(i)-d(i) | = \ X (n-1)+Iz(n)-d(n)|2 (5), i=1 where is the omission factor of the algorithm (0 < < 1). The stationary environment corresponds to an omission factor equal to 1. In order to follow channel variations as efficiently as possible, the algorithm must minimise t (n) for each sample n of signal d(t), which means that it is necessary to JO know the response d(n) to each sample. The response is by definition only known on learning sequences. On information symbol sequences, it is possible to continue the algorithm's adaptation by using the principle implemented in the DEE equaliser, which first calculates the output z(n) obtained using the filters optimised at moment n-1 and then determines the :: symbol d(n) . The symbol d(n) thus estimated is used as a response, d(n)=d(n), to carry out a new iteration of the algorithm.

For each sample n the signal z(n), on which decision module 18 works, is divided into three quantities derived respectively from the spatial part, the recursive part and the transverse part. Signal z(n) is therefore defined by the JO following formula:

1 5 z(n) ZS(n)-ZR(n)+ZT(n) (6) At the input of the spatial part, the following signal vectors, each one associated with one of sensors On, where n = 1 to N. are used by the adaptive algorithm and have the following form: 5 Xk(n)=X(nTs+pkTe) fork=1,, K (7) i.e. Xk(n)=ockd(nTs)Sk+ id[nTs-(pi-pk)Te]si+B(nTs+pkTe) (8), ink where k corresponds to a determined path selected at synchronization stage 1. lo Thus, each one of these vectors Xk(n) contains a part correlated with the response d(n), the term ockd(nTs) Sk and an ISI part which must be compensated for by the spatial part and/or the temporal part. The adaptive algorithm will seek to realign the phase of the different contributions of vectors Xk(n) correlated with response d(n).

Is The advantage of the proposed structure is therefore clearly evident upon analysis of formula (8): to phase realign the K paths arriving at the antenna, i.e. to "take advantage of,' the energy of the K paths in the equaliser, it is not necessary to place a FPR filter on each sensor as in the multi-sensor equaliser proposed So by Scott et al. The dimensions of the FPR filter must be linked to the size of the channel and therefore contain a large number K' of coefficients. All that is required is to insert one filter containing K coefficients, which is equivalent to selecting K coefficients from the K' coefficients making up the FPR filter of the multi-sensor equaliser proposed by Scott et al. The number of coefficients is as therefore greatly reduced, which means that the adaptive algorithm can converge towards the optimum solution more quickly. Given that the algorithm adapts to the learning sequences, i.e. to a given number of iterations, the proposed structure therefore produces improved performance in comparison with the multi-sensor equaliser proposed by Scott et al. In addition, in order to so guarantee good results in non-stationary environments, the number of coefficients to be adapted should be reduced as far as possible.

The output of the spatial part is expressed by the following formula, the weight vector weighting signal Xk(n) being represented by Wk:

K zS(n)=Y(n)= 2 wk+xk(n) k=1 where + in superscript represents the transposition-conjugation operation.

By using the notation Xs(n)=[XT(n)...XKT(n)]T, where T in superscript represents the transposition operation in a vectorial space, to represent the s input signal vector of the spatial part and Ws=[W1T..WKT] T to represent the weighting vector of the spatial part, the output signal of the spatial part is expressed by the following formula: ZS (n) = y(n) = Ws+Xs (n) (10).

The output of the recursive part is written as a function of HR, the filter lo weighting the recursive part, and of symbols n-1 to n-R, and is expressed by the following formula: R zip (n) = 2, H Rj d(n - i) (1 1), i=1 A where d(n) = d(n) on the information sequences, and where * in superscript represents the conjugation operation on complex 5 numbers.

Symbols n-1 to n-R are either the known symbols in the learning sequences or the symbols, d(n) = d(n), determined during the previous iterations on the information sequences.

The input samples of the transverse part are calculated by calculation do unit 16 on the basis of the signals output by unit 13 and from filters Wk at instants n+1 to n+T, and are therefore dependent upon the weighting system of the spatial part. Two methods can be used in the algorithm for the calculation of these samples: - a first method consists in updating all the samples of the transverse 95 part with the vector Ws(n-1) calculated at the previous iteration using the following formula: y(n+kin1)=ws(n-1)+xs(n+k) k=1,...,T (12) A second method can be used which enables the optimization of the calculation power:

- under this second method the transverse part consists of a delay line.

For the symbol n, the algorithm therefore calculates only the sample y(n+ T/n-

1) on the basis of the weighting vector Ws(n-l), the other samples having been calculated during the previous iterations y(n+T-1/n-1) is therefore 5 calculated on the basis of weighting vector Ws(n-2), y(n-T-2/n-1) on the basis of Ws(n-3) and so on.

On the basis of samples y(n+k/n-1) calculated using one of the above methods, the output of the transverse part is expressed by the following formula: in ZT(n)=iHTi y(n+i/n-1) (13), i=1 where HT is the filter of the transverse part.

At each iteration of the algorithm for the update of the system of filters (W. HR, HT) making up the system, the samples of the transverse period y(n+k/n-1) for k = 1,..., T must be calculated first. The samples thus 15 calculated become the input of the adaptation algorithm in the same way as the vector X(n) and the symbols corresponding to the previous iterations d(n-1) to d(n-R). The adaptation algorithm then seeks the system (W(n), HR(n), HT(n)) which minimises the criterion t (n).

Different algorithms can be used to calculate the filter system (W. HR, so HT) giving the minimum value of estimated MOE t (n) for each iteration. The algorithm selected is a least squares algorithm, which is chosen in preference to a gradient algorithm for reasons of speed of convergence. Of the least squares algorithms available, the spatial matrix algorithm illustrated in figure 6 is used to provide joint adaptation of the spatial and temporal part. Any other as least squares algorithm would produce the same basic results.

The spatial matrix algorithm does not estimate the filter system (W.

HR, HT) directly. At each iteration the samples corresponding to the spatial, recursive and transverse parts are fed into the matrix structure and the coefficients of the matrix C)(i, j), which are also termed adaptive multipliers, so are calculated so as to minimise the power of the estimation error e(n) = z(n)-d(n). The spatial matrix algorithm exists in two versions: the "a priori" version and the "a posterior)" version. The "a priori" version is

expressed by the following series of instructions, based on the one hand on the known sequence symbols; the order of the matrix is noted: order = R+T+S+1.

- in an initiaiisation phase: 5 i=1->R E(i)=d(n-i) initialization of the recursive part i=1 ->T E(R+i)=y(n+i/n-1) initiaiisation of the transverse part i=1->S E(R+T+i)=X(i) initialization of the spatial part, where x(i) is the ith component of vector X(n) i = Order E(Order)=d(n) initialization of the response signal lo - then from p = 1->Order: a(p) = la(p) + 7(P 1)||E(P)||2 T(P) = 7(P -1) - T(P -1) E(p) / a(p) i=p+1 -> Order: E(i)=E(i)-C(i,P) E(p) C(i,p)=C(i,p)+ (p-1)E(P)E(i) /(x(p) and on the other hand on the information symbol sequences: the initiaiisation phase is identical to the preceding initiaiisation phase, apart from the fact that E(Order) is not initialized as the response is not known. The response must therefore be estimated. To do this, the adaptation go algorithm first updates the various quantities involved in the spatial matrix algorithm which have no effect upon E(Order), i.e.: - from p=1-> 0rder-1: (I) = (I) + T(P - 1)||E(P)||

)(P) = T(P - 1) + 7(p - 1)2||E(P)|| / (up) :5 then from i=p+1 Corders: E(i)=E(i)-C(i,P) E(p) C(i,p)=C(i,p)+ (p- 1)E(p)E(i) /ot(p) The output of the multi-sensor equaliser is then calculated using the different error signals and is expressed by the following formula: 30 z(n) = C(Order,i) E(i) (14).

i=] Decision module 18 then determines signal d(n) on the basis of z(n) and updates the final part of the matrix structure:

1 9 - in an initiaiisation phase: E (O rde r) =d (n) then from p=1 Order-1: E(Order) = E(Order)-C(Order,p)*E(p) 5 C(Order, p) = C(Order, P) + (p-1)E(p)E(Order)*/oc(p) For the calculation of the transverse part at each iteration, the samples of the transverse part, i.e. the samples obtained at the output of the spatial part for X(n+1),..., X(n+T), must be calculated first.

The output of the spatial part corresponds to the contribution of the lo spatial part to the signal subtracted from the response. The signal subtracted from the response is expressed on the basis of the different error signals E(1)->E(Order-1) by the following formula: z = C(Order,i) * E(i) (15).

In order to calculate the output of the spatial part corresponding to Is X(n+k) where k=1->T, it is therefore necessary to simply calculate the contribution of the spatial part to the different error signals. In order to reduce the calculation power, the samples corresponding to the spatial part are placed on the right of the matrix, and it is therefore only necessary to calculate the contribution of the spatial part to error signals (E(R+T+1) to do E(R+T+S): this means that only that part of the matrix marked by the solid line in figure 6 is involved in the calculation.

Let us suppose that Es(i) is the contribution of the spatial part to error signal E(i). Es(i) is thus calculated using the following series of instructions: - in an initialization phase: 25 Es(i+R+T)=x(i) for i + 1-> S, where x(i) is the ith component of vector X(n+k).

- then for 1 = R+T+1->R+T+S, update of error signals Es(j) for j = i+1->R+ T+S on the basis of error signal Es(i): i = R+T+1->R+T+S 30 j = i+1 ->R+T+ S ES(i)=Es(i)-C(J,i)*ES(i) The output of the spatial part is thus expressed as a function of error signals Es(i) calculated above by the following formula: y = C(Order,i) * E(i) (Order-1 =R+T+S) (16).

The following example demonstrates the relative usefulness of the spatial and temporal parts of the multi-sensor equaliser.

A useful signal arrives at the antenna along two propagation paths.

The signal vector received by the antenna is expressed by the following 5 formula: X(t) = cr1d(t)S1 + a2d(t - T)S2 + B(t) (17).

According to formula (7), the spatial part of the structure is composed of the vectors X(t) and X(T+t). The output of the spatial part y(t) is therefore expressed as follows: y(t) = W1 X(t) + W2 X(t + 1) (18) That is: y(t) = d(t)[Cl1 W1 S1 + 2 W2 S2] + d(t-1)[0 2 W1 S2 +d(t+T)[(x1W2 S1]+ (19) W1 B(t) + W2 B(t + 7) The output of the spatial part therefore consists of three components: one component corresponding to the useful signal d(t), one component 15 corresponding to the ISI generated by d(t-l) and d(t+), and one component corresponding to the noise (background noise + interference).

Let us now suppose that the temporal part of the structure is absent: T=R=O. The algorithm adapting the structure minimises the MOE between y(t) and d(t), and therefore seeks to cancel the two terms containing ISI, since do these are non-correlated with respect to response d(t). ISI is processed by the antenna in the same way as any possible interference.

The following simulation is designed to analyse the behaviour of such a structure without a temporal part. An example antenna used for the simulation featuring five monidirectional sensors C1 to Cs disposed on the sides of an : equilateral triangle is illustrated in figure 7. The angle between any two sides of the triangle is 60 .

The antenna receives two non-correlated paths of identical bearing of 0 and of power s = 10 dB. The elevation of the first path is 40 and that of the second path is varied. The background noise has a power of C72 = 0 dB.

The output powers of the useful signal, the ISI and the background noise are

expressed respectively by the formulas given below: S = |a1 W1 S1 + a2 W2 S2| = S[W1 S1 + W2 S2] (20) IIS= S lW1 S2| +|W2+S1l (21) 5 B = W1 Rbb W1 + W2 Rbb W2 = a2 [W1 W1 + W2 W2] (22) In figure 8, the curves S/(ISI+B), S/B, iSI/B are traced on a Cartesian co-ordinate graph, on which the yaxis represents the amplitude in dB and the x-axis represents the elevation angle in degrees. The antenna processes ISI in the same way as interference, and therefore optimises the ratio S/(ISI+B).

lo When the coefficient of spatial correlation between the two paths is low, the ratio S/(ISI+B) at the output of the antenna is close to 20 dB. The determination of the symbols emitted is carried out on signal y(t), which means that the same performance as on a stationary channel featuring one path of power 20 dB is obtained. In comparison, the single- sensor DFE equaliser produces performances similar to those of a stationary channel featuring one path of power 10dB. The processing carried out therefore enables a gain of 10 dB, a gain which is made up as follows: 7 dB = 10 logN due to the gain in S/B of the antenna aligned in the direction of each of the two paths.

o - 3 dB due to the phase realignment gain of the two paths.

In such a configuration, the spatial part eliminates ISI, directs one lobe in the direction of each of the two paths and phase realigns the two paths.

The temporal part no longer makes any useful contribution.

When the two paths are close in spatial terms, it becomes more and 5 more difficult for the antenna to eliminate ISI while at the same time maintaining a S/B gain which is sufficient for the two paths.

Thus, for elevations c 36 or > 44 , the antenna is always able to reject ISI below the level of the background noise, but this is achieved at the cost of

a deterioration of the S/(ISI+B) ratio with respect to the previous case (at 36 , o 12 dB are lost). The performance of the decision module is therefore lower than in the previous case.

For elevations between approximately 36 and 44 the two paths are too close in spatial terms, and the antenna is no longer capable of eliminating ISI. The ratio S/(ISI+B) tends towards 3 dB. It should be noted that this result is obtained whatever the value of the common power of the two paths. A 5 decision module placed at the output of the spatial part would therefore give less satisfactory results than the singlesensor DFE equaliser, which is evidently unacceptable.

The disadvantage of such a structure without a temporal part is therefore clear from the analysis of this example: ISI is processed by the lo antenna in the same way as interference, and the antenna therefore uses varying degrees of liberty to eliminate it.

The addition of a temporal part to the structure and the adaptation of the temporal and spatial parts to the same error signal results in the overall behaviour described below.

15 For paths which are "sufficiently" non-correlated in spatial terms, the spatial part enables the system in all cases to direct the main lobe of the antenna in the direction of each of the two paths and to realign their phase, while at the same time eliminating ISI. The work of the temporal part is therefore reduced. The overall gain with respect to the single-sensor DFE no equaliser is 10 log N+3 dB.

For spatially correlated paths the temporal part handles the elimination of ISI, which means that the spatial part will no longer seek to optimise the antenna gain in the direction of each of the two paths. The overall gain with respect to the single-sensor DFE equaliser is 10 log N. 95 A gain with respect to a single-sensor DFE equaliser of between 10 log N and 10 log N+3dB for two paths of the same power in a stationary environment can therefore be realised in all cases.

Moreover, in the presence of jamming of a useful signal, the spatial part will reject interference.

Claims

C L A I M S

1. A pro for nulti or eq isatia in a radio receiver ca risiry; 5 a spatial part connected to a temporal part composed respectively of a determined number of filters and receiving a radio signal consisting of at least a learning sequence made up of symbols known to the receiver and an information sequence made up of useful symbols, and comprising a first stage for the preliminary processing of the signal received by the receiver, lo of the transformation of the signal received by at least two sensors into an equivalent baseband signal, of the sampling of the baseband signal at a rate which is a multiple of the symbol rate and of the filtering of the sampled signal using a low-pass filtering process, wherein, in the presence of interference and multiple propagation paths, it consists, in 15 order to reduce the number of filter coefficients to be adapted, a second synchronization stage, of a synchronization measurement process, of the estimation of the number of paths in the signal and the delay times associated with the various paths and their relative powers, and of the estimation of the frequency offset between the emission and reception of JO the signal in order to compensate for it, and wherein, in a third multi-sensor equalization stage, it consists in selecting a determined number of paths according to a determined criterion from the number of paths estimated in the synchronization stage, in filtering via a spatial processing procedure the signal received by the receiver using the filters of the 25 spatial part, in filtering via a temporal processing procedure the signal output by the spatial part using the filters of the temporal part, the respective coefficients of the filters of the spatial part and the temporal part being jointly and periodically recalculated at each iteration by an adaptive algorithm working at the symbol rate in order to 30 minimise the estimation error produced between the receiver output signal and the response signal.

2. A process accor ir to n3; i n i, to each path selected at each iteration a signal vector, end comprising

filtering in the spatial processing stage each signal vector using a filter of the spatial part, and calculating on the basis of the result of the filtering of the previous iteration and the paths selected at the current iteration the input signals of the transverse part of the temporal 5 part of the equalization.

3. A process accor lir4; to claim 1 or claim 2, apprising the to processing stage filtering in the transverse part of the temporal part the signals output by the filters of the spatial part, summing the output lo signal of the spatial part with the output signal of the transverse part, and subtracting the signal derived from the recursive part of the temporal part from this sum, the recursive part using a filter to filter the symbols of the previous iterations, which are the symbols "determined" on the basis of the information sequences or the known symbols of the learning sequences, in 15 order to deduce the output signal of the receiver.

4 A process ac rdi to claim 1, In the criteria capsizes selecting a maximum determined number of paths in order to limit the number of coefficients to be calculated in the spatial part.

5. A pros according to Him 1, Rein He criteria Rises selecting the paths whose relative power with respect to a main path is greater than a determined threshold in order to limit the number of coefficients to be calculated in the spatial part.

6. A pr ess aa ir g to Rim 1, Therein He criteria at using selecting simultaneously a maximum determined number of paths and the paths whose relative power with respect to a main path is greater than a determined threshold in order to limit the number of coefficients to be so calculated in the spatial part.

7. A Recess accordir to my ore of come 1 to 6, In the adaptive algorithm cat musing Mean Quadratic Error (MQE) between the receiver output signal and a response signal composed of

known symbols in the learning sequences and determined symbols in the information sequences by weighting the MOE samples at the symbol rate.

8. A process according to any one of cluing 1 to 7, wherein the adaptive algorithm is a spatial matrix algorithm.

9. A radio receiver featuring at least one nn ti-sensor spat-i 1 diversity 5 equaliser cases a spatial part connected to a temporal part and receiving a digital radio signal composed of at least a learning sequence made up of symbols known to the receiver and an information sequence made up of useful symbols, featuring, in order to reduce the number of filter coefficients to be adapted in the spatial and temporal part in the presence of to interference and multiple propagation paths: - at least two sensors connected to a unit carrying out the preliminary processing and synchronization of the receiver input signal, the outputs of the unit being connected respectively to a first series of inputs and a second series of inputs of the spatial part of the equaliser, the < first series of inputs corresponding respectively to the inputs of the spatial filters relating to each of the paths selected from a determined number of paths detected, and the second series of inputs corresponding respectively to the inputs of a unit for the calculation of the input signals of the transverse part of the temporal part of the equaliser, and wherein the ho transverse part of the temporal part features a transverse filter of determined coefficients, the temporal part also featuring a recursive part consisting of a decision module whose output is connected to the input of a recursive filter of determined coefficients, the recursive filter (HR) being located in a loop and receiving on its input the sum of the output 25 signals of the spatial part and of the transverse part, from which is subtracted the signal output by the recursive filter 10. A process for nulti-ffensor equalization in a radio receiver, substant.i lly as described hereinbefore with reference to the acoc p!nyung drawings 11. A radio receiver featuring at least one multi- sensor spatial diversity equalizer substantially as described hereinhefore, with reference to the Attica paoyuqe drawings, and as illustrated in Figure 5 of these drawings.

Or Amendments to the claims have been filed as follows 1. A process for multi-sensor equalization in a radio receiver consisting of a spatial part connected to a temporal part composed respectively of a determined number of adaptative filters and receiving a radio signal consisting of at least a learning sequence made up of symbols knovvn to the receiver and an information sequence made up of useful symbols comprising the steps of: transforming the signal received by at least two sensors 10 into an equivalent base band signal, - sampling the base band signal at a rate which is a multiple of the symbol rate - filtering the sampled signal using a low pass filtering process 5 - estimating the number of paths of the signal by a synchronization measurement process - selecting a determined number of paths according to a determined criterium from the number of estimated paths - filtering via a spatial processing procedure the signal 2Q received by the receiver using adaptative filters of the spatial part - filtering via a temporal processing procedure the signal output by the spatial Dart using adaprative filters of the temporal part combining the spatial part output and the temporal part output to 25 produce a receiver output - comparing the receiver output with either the learning sequence or the information sequence to provide an input to the adaptative processing for periodically recalculating at the symbols rate the respective coefficients of the adaptative filters of the spatial part 30 and the temporal part in order to minimise the estimation error produced betvveen the receiver output signal and either the learning sequence or,information sequence.

2. A pros acmnlir to claim 1, herein tore Arks to each path selected at each iteration a signal vector. andcomprising

erg filtering in the spatial processing stage each signal vector using a filter of the spatial part, and calculating on the basis of the result of the filtering of the previous iteration and the paths selected at the current iteration the input signals of the transverse part of the temporal 5 part of the equaiisation.

3. A process acorn to chum 1 or chum 2, risir the temporal processing stage filtering in the transverse part of the temporal part the signals output by the filters of the spatial part, summing the output lo signal of the spatial part with the output signal of the transverse part, and subtracting the signal derived from the recursive part of the temporal part from this sum, the recursive part using a filter to filter the symbols of the previous iterations, which are the symbols "determined" on the basis of the information sequences or the known symbols of the learning sequences, in 15 order to deduce the output signal of the receiver.

4 A process accordir to claim 1, In to} criteria c rises selecting a maximum determined number of paths in order to limit the number of coefficients to be calculated in the spatial part.

JO 5. A Recess arx rdir to chin 1, Rein the criterion ca r selecting the paths whose relative power with respect to a main path is greater than a determined threshold in order to limit the number of coefficients to be calculated in the spatial part.

6. A price acquire to chin 1, herein the criteria chrism using selecting simultaneously a maximum determined number of paths and the paths whose relative power with respect to a main path is greater than a determined threshold in order to limit the number of coefficients to be JO calculated in the spatial part.

7. A prices acoordi to are of cub; 1 to 6, herein the adaptive algorithm campuses onnirasing Mean Quadratic Error (MQE) between the receiver output signal and a response signal composed of

I known symbols in the learning sequences and determined symbols in the information sequences by weighting the MOE samples at the symbol rate.

8. A As at to any cre of coin 1 to 7, Herein Me adaptive algorithm is a spatial matrix algorithm.

9. A radio receiver operable to carry cat a ^ a rdir to said As aE t al this i ticn i: Fens for to sigra1 roil}fly at OF - the ours iITtO al a iva t nal, Arms far sample 1 Final at a rate Chide is a multiple of the sow rate, As far filterdrg the sitars a lo Fax; filterdrg Is, ns for estimate of the row of paths of the signal lay a s i tia not;, reels fcr se tirO a do row of Fates sordid to a te rdrf c:riterinn Fran to For cf estimate Atlas, mess fcr filbarirO via a spatial Rosin he the signal remit by the rEnei r use ac ptati filters Of spatial At, beans for filteeirO via a ter EaL pi prr e the signal at fly Off SEz.i 1 Far; annotative filler Of t aL part, An for rbinirg spatial part Q=t are tonal part abut to Fog a rosier aft, rEa3nS far c rirg the raider Q-t Edith either the levity ED ar the infor-

gratis set to pelvis an irrupt to the ac ptati e Dig fcr Ferictli y r aGatim at 1 rate the retire fficib L of the a iu filters cf the sF i ' F-

arb Off tarsal Fart in ardor to ironic the mbirratia ever 1 the rem eiver Ant signal am either bemire He cr induration so.

10. A Is fa" ti r-=li ticn in a Rio receiver, tanti lly as described reirief With refererae to to at firm.

11. A radio Reviver. featuring at loot cue lti 5mli 1 diversity - =li a tar ally as descritecl tereirt efcs e, with referee to the afire chains, art as illustrated in Fit 5 of the drswir -.