JPH09205392A - Radio communication system and radio receiver - Google Patents

Abstract

PROBLEM TO BE SOLVED: To eliminate an undesired wave and to reduce an error rate of the receiver by synthesizing a direct wave and a 1st delay signal resulting from a received radio signal and providing an output of the synthesized wave as the reception signal. SOLUTION: An adaptive control processor 9 of a receiver 30 samples a received signal at a sampling period Ts equal to the time length of a symbol string from a reference time being a time of a 2nd symbol in a 1st symbol rate of a training signal included in the desired wave of the received signal at first to recover a recovered training signal. Then the weight to each antenna element 100-i is calculated and outputted to a beam shaping device B1. Then each multiplier 2-i (i=1, 2,..., N) in the beam shaping device B1 multiplies a reception signal xi (t) with a complex weight wi and provides an output of the product signal to an adder 3 and the adder 3 adds N-sets of signals outputted from the multipliers 2-1-2-N to provide an output of a signal y(t).

Description

Detailed Description of the Invention

[0001]

The present invention relates to a radio communication system and a radio receiver used in the radio communication system.

[0002]

2. Description of the Related Art In high-speed digital mobile communication, an adaptive array antenna is used to improve frequency selective fading. The adaptive array antenna suppresses the multi-propagation wave by forming a directional null in the direction of the multi-propagation wave to reduce frequency selective fading. Recently, after separating a received signal into a direct wave signal and a delayed wave signal,
2. Description of the Related Art A desired wave is received by effectively using the power of each delayed wave by correcting and combining delay time differences.

A conventional receiver is disclosed in, for example, prior art document 1, "Yasutaka Ogawa, et al.," Study of spatial domain path diversity using adaptive array ", IEICE General Conference,
pp 397, 1995 ". The receiver of the related art obtains a desired signal by separating and extracting the direct wave signal and each delay wave signal, correcting the delay time difference, and combining the signals.

[0004]

However, in the conventional receiver, when the correlation between the direct wave signal separated from the received signal and the delayed wave signal and the correlation between the respective delayed wave signals are strong, the output signal is low. However, there is a problem that the signal-to-noise ratio cannot be increased. Further, if the delay time of the incoming delayed wave is not known, the direct wave signal and each delayed wave signal cannot be properly separated, so that the signal-to-interference wave ratio of the output signal cannot be increased. there were. As a result, there is a problem that the error rate of the received signal cannot be reduced.

A first object of the present invention is to solve the above-mentioned problems and to appropriately separate a direct wave signal and each delayed wave signal, and to reduce the number of receivers compared with a conventional example. An object of the present invention is to provide a wireless communication system capable of reducing an error rate.

A second object of the present invention is to solve the above problems and to appropriately separate a direct wave signal and each delayed wave signal, and to reduce the error rate as compared with the conventional example. An object of the present invention is to provide a wireless receiver that can be reduced in size.

[0007]

According to the first aspect of the present invention, there is provided a wireless communication system including a transmitting means for transmitting a wireless signal including a predetermined first training signal and a subsequent data signal. A wireless transmitter, and receiving means for receiving a wireless signal transmitted from the wireless transmitter based on a first training signal included in the wireless signal and a predetermined second training signal. A wireless communication system including a wireless receiver, wherein the first training signal is formed by serially serializing a plurality of Q sets of symbol sequences each including a plurality of P identical symbols; The training signal is the multiple Q
One symbol of each set of the set of symbol sequences is serially continuous, and the wireless receiver includes the wireless signal received by the receiving unit in the first received direct wave wireless signal. A time after a predetermined second symbol of the first symbol sequence of one training signal as a reference time, the reference time and a sampling interval equal to the time length of each symbol sequence from the reference time. Based on a correlation between the second training signal and the third training signal, and sampling means for outputting a third training signal composed of a sequence of symbols sampled and sampled at
The radio signal received by the receiving means is divided into the direct wave and a first signal which arrives first after being delayed from the direct wave signal.
And a signal control means for combining at least the delayed wave signal and outputting as a received signal.

According to a second aspect of the present invention, there is provided the wireless communication system according to the first aspect, wherein the reference time is determined based on the first training signal included in the first received direct wave wireless signal. Is the time of the P-th symbol in the symbol sequence.

Further, a wireless communication system according to claim 3 is the wireless communication system according to claim 1 or 2,
The wireless receiver includes an array antenna composed of a plurality of N antenna elements juxtaposed and arranged in a predetermined arrangement shape, and the signal control unit includes a third training signal and a second training signal. For weighting the third training signal so as to synthesize only a signal that matches the second training signal based on the correlation vector between the second training signal and the autocorrelation matrix of the third training signal. Adaptive control means for calculating and outputting an N-dimensional weight vector; and multiplying the weight vector by an N-dimensional column vector having N received signals received by each of the antenna elements as components. Beam forming means for combining a signal including the direct wave signal and the at least one delayed wave signal and outputting the combined signal as a received signal. It is characterized in.

According to a fourth aspect of the present invention, there is provided a radio receiver comprising: a radio signal transmitted from a radio transmitter for transmitting a radio signal including a predetermined first training signal and a subsequent data signal; The first signal included in the radio signal
And a receiving means based on a predetermined training signal and a predetermined second training signal, wherein the first training signal is composed of a plurality P of the same symbols. The plurality of Q sets of symbol sequences are serially continuous, and the second training signal is such that one symbol of each set of the plurality of Q sets of symbol sequences is serially continuous, and the radio receiver includes:
The radio signal received by the receiving means is converted to a time after a predetermined at least a second symbol of the first symbol sequence of the first training signal included in the first received direct wave radio signal. Sampling means for setting the reference time as the reference time and outputting a third training signal composed of a sequence of symbols sampled by sampling at a sampling interval equal to the time length of each symbol sequence from the reference time, On the basis of the correlation between the second training signal and the third training signal, the radio signal received by the receiving means is converted to the direct wave and the first signal that arrives first after being delayed from the direct wave signal. And a signal control means for combining at least the delayed wave signal and outputting as a received signal.

According to a fifth aspect of the present invention, in the wireless receiver according to the fourth aspect, the reference time is equal to the first time of the first training signal included in the first received direct wave radio signal. Is the time of the P-th symbol in the symbol sequence.

The radio receiver according to a sixth aspect of the present invention is the radio receiver according to the fourth or fifth aspect, further comprising an array antenna comprising a plurality of N antenna elements juxtaposed and arranged in a predetermined arrangement shape. Wherein the signal control means comprises:
Based on a correlation vector between the third training signal and the second training signal and an autocorrelation matrix of the third training signal, the third training signal is Adaptive control means for calculating and outputting an N-dimensional weight vector for weighting so that only matching signals are synthesized; and an N-dimensional signal having N received signals received by each of the antenna elements as components Beam forming means for multiplying the column vector by the weight vector to synthesize a signal including the direct wave signal and the at least one delayed wave signal and outputting the combined signal as a received signal. And

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS A communication system according to an embodiment of the present invention will be described below with reference to the drawings. The communication system according to the present embodiment includes a transmitter 50 and a receiver 30, and has the following features. (1) As shown in FIG. 3, the transmitter 50 transmits a first symbol sequence composed of two identical symbols A, a second symbol sequence composed of two identical symbols B, A third symbol sequence of the same symbols C;
A training signal in which a fourth symbol sequence composed of two identical symbols D and a fifth symbol sequence composed of two identical symbols E are serially continuous;
A transmission signal including the data signal is transmitted. (2) In the receiver 30, the signal generator 10 generates a reference training signal d (t) having a reference symbol sequence [A, B, C, D, E]. (3) In the receiver 30, the adaptive control processor 9
As shown in FIG. 4, it is equal to the time length of the symbol sequence with reference to the time of the second symbol of the first symbol sequence of the training signal included in the transmission signal of the desired wave received first. The reproduction training signal is sampled at the sampling cycle Ts to reproduce the reproduction training signal, and the weight vector W is calculated based on the reproduction training signal and the reference training signal d (t) and output to the beam former B1. (4) In the receiver 30, the beamformer B1 forms a beam on the basis of the weight vector W calculated by the adaptive control processor 9, and outputs a desired wave arriving from a desired transmitter 50 and one from the desired wave. A delay wave arriving with a delay time within the symbol time is output.

Hereinafter, a communication system according to an embodiment will be described in detail with reference to the drawings. First, in the communication system according to the embodiment, the transmitter 50, as shown in FIG.
Training signal insertion circuit 51, training signal generation circuit 52, modulator 53, intermediate frequency signal generator 54, frequency conversion circuit 55, power amplifier 56, transmission antenna 57
Consists of In the transmitter 50, the training signal generation circuit 52 includes a predetermined signal, as shown in FIG.
The training signal is generated by serially connecting five sets of symbol strings each including the same symbol, and outputs the training signal to the training signal insertion circuit 51. The training signal insertion circuit 51 inserts the training signal at the head of the input baseband data signal, and outputs the baseband signal with the training signal inserted to the modulator 53. Intermediate frequency signal generator 54
Generates an intermediate frequency signal having a predetermined intermediate frequency, and outputs the intermediate frequency signal to the modulator 53. The modulator 53 modulates the input intermediate frequency signal according to a baseband signal into which the input training signal is inserted, for example, using a modulation scheme such as PSK, and converts the modulated signal into a frequency conversion circuit. Output to 55. Here, the present invention is not limited to PSK, but PSK, BPSK or Q
Other modulation schemes such as AM may be used. The frequency conversion circuit 55 includes a local oscillation signal generator, a mixer, and a low-pass filter, and performs low-pass filtering after mixing the input modulated signal and the input local oscillation signal. The frequency of the radio signal is converted to a radio signal of a predetermined radio frequency, and the radio signal after the frequency conversion is
7 to output as a transmission signal. The transmitter 50 configured as described above transmits a transmission signal that is a radio signal including a training signal and a baseband data signal.

Here, in the communication system of the embodiment, the training signal included in the transmission signal transmitted by the transmitter 50 is, as shown in FIG. 3, a first symbol sequence consisting of two identical symbols A. , A second symbol sequence composed of two identical symbols B, a third symbol sequence composed of two identical symbols C, and a fourth symbol sequence composed of two identical symbols D The column and a fifth symbol column composed of two identical symbols E are serially continuous.

In the communication system according to the present embodiment, as shown in FIG. 1, a receiver 30 includes a predetermined plurality of N antenna elements 100-1 arranged in close proximity in a predetermined arrangement shape.
Antenna 100, which is composed of a plurality of array antennas 100 to 100-N, receiving modules RM-1 to RM-N, and an A / D converter AD
-1 to AD-N, in-phase distributors 1-1 to 1-N,
It comprises a signal controller 20, a maximum likelihood estimator 7, and a demodulator 8. Here, the signal controller 20 includes a beam former B1.
, Adaptive control processor 9, signal generator 10, RO
M11.

Hereinafter, the receiver 30 in the communication system according to the embodiment of the present invention will be described in detail with reference to FIG.

In the receiver 30 of the embodiment, the receiving module RM-i includes a low-noise amplifier (not shown) and a down-converter (not shown).
The received signal received at 00-i is amplified by a low-noise amplifier, frequency-converted by a down-converter into an intermediate frequency signal of a predetermined intermediate frequency, and the frequency-converted intermediate frequency signal is converted to an A / D converter AD. Output to -i. The A / D converter AD-i samples the frequency-converted intermediate frequency signal at a sampling period for each symbol of the transmission signal, performs A / D conversion, and receives the received signal Ri, which is a digital signal after A / D conversion. To the in-phase distributor 1-i.

The common-mode distributor 1-i includes a receiving module RM
-I, the received signal Ri output from the two received signals x
_i (t) and output to the multiplier 2-i and the adaptive control processor 9. Here, i in the code is a code representing the code 1 to N, that is, i = 1,
2,..., N, and in the present specification, the same applies unless otherwise specified.

The signal generator 10 responds to a command signal input from the adaptive control processor 9 described later, and
1, a reference training signal d (t) is repeatedly generated with a symbol period T equal to the time length of each of the first to fifth symbol sequences of the training signal included in the transmission signal. And outputs it to the adaptive control processor 9. Here, the adaptive control processor 9
When the _{first one} of the received signals x ₁ (t) to x _N (t) is input, a command signal for commanding the signal generator 10 to start generating the reference training signal d (t) is issued. Output. In the ROM 11, a reference symbol sequence [A, B, C, D, E] composed of one symbol of each of the five symbol sequences of the training signal included in the transmission signal is stored in advance. That is, in the reference training signal d (t) generated by the signal generator 10, one symbol of each set of five sets of symbol sequences of the training signal included in the transmission signal is serially continuous.

As shown in FIG. 4, the adaptive control processor 9 determines two of the first first symbol sequence of the training signal included in the transmission signal of the desired wave received first.
Using the time of the second symbol as a reference time, the reproduction training signal is reproduced by sampling the reference time and the reference time at a sampling period Ts equal to the time length of the symbol sequence. Then, based on the reproduction training signal and the reference training signal d (t), the adaptive control processor 9 performs MMSE (Mini), which is the least square method, during the training signal period shown in FIG.
The weight of each antenna element 100-i is calculated based on a method described later, which is called a MM Mean Square Error (algorithm), and the calculated weight is output to the beamformer B1.

Here, the arithmetic processing of the adaptive control processor 9 based on the MMSE algorithm will be described. First, an N-dimensional column vector X (t) is defined by the following equation _{1 using} the received signals x ₁ (t) to x _N (t), and the signal y (t) to be output by the beamformer B1 ) Using the weight vector W and the N-dimensional column vector X (t),
Expressed by Here, the weight vector W is a vector having N complex weights w _{1 to} w _N as components.

[0023]

X (t) = [x ₁ (t), x ₂ (t),..., X _N (t)] ^T

## EQU2 ## y (t) = W ^{T ×} X (t)

Here, [] ^T in Equation 2 represents the transposition of a matrix.

In the communication system of the present embodiment, as described above, the transmitter 50 transmits the first symbol sequence composed of two identical symbols A and the second symbol sequence composed of two identical symbols B. , A third symbol sequence composed of two identical symbols C, a fourth symbol sequence composed of two identical symbols D, and a fifth symbol sequence composed of two identical symbols E A transmission signal including a training signal in which the symbol sequence is serially continuous is transmitted.

In the receiver 30, the adaptive control processor 9 uses the time of the second symbol of the first symbol string of the training signal included in the transmission signal of the desired wave which is first received as a reference, as described above. At time, the reproduction training signal is reproduced by sampling the received signal at the reference time and the sampling period Ts equal to the time length of the symbol string from the reference time.
Thereby, as shown in FIG. 4, for example, the reproduction training signal included in the signal corresponding to the desired wave arriving from the transmitting station earliest and the first delayed wave arriving after one symbol time from the desired wave is: One symbol of each set of the five sets of symbol sequences of the training signal included in the transmission signal is serially continuous. Therefore, when the radio channel is good, the reproduction training signal included in the signals corresponding to the desired wave and the delayed wave is the reference training signal d.
It has the same symbol sequence [A, B, C, D, E] as the reference symbol sequence [A, B, C, D, E] of (t). That is, by calculating the weight vector W such that the waveform of the synthesized signal is as similar as possible to the waveform of the reference training signal d (t) during the training signal period, the synthesized signal including only the desired wave and the first delayed wave is synthesized. Weight vector W for performing the calculation. In other words, the adaptive control processor 9 weights the weight vector W for weighting so that only the signal whose reproduced training signal matches the reference training signal d (t) is synthesized.
Is calculated and output.

Here, the waveform of the signal y (t) to be output during the training signal period and the reference training signal d
An error signal ε indicating the degree of similarity to the waveform of (t) is defined as in the following Expression 3. That is, the instantaneous square error ε ² (k = 1,2) obtained by squaring the error signal ε defined by the following equation (3)
2,..., L) is minimized.
Is calculated, a weight vector W for extracting a combined signal including only the desired wave and the first delayed wave can be obtained. Here, the second expression on the right side in Expression 3 is obtained by substituting the expression on the right side of Expression 2 for y (t) in the first expression on the right side. Also, by squaring both sides of Equation 3, the instantaneous square error ε ² can be expressed by Equation 4, and the mean square error E [ε ² ] can be expressed by the following Equation 5. it can.

[0028]

Ε = d (t) −y (t) = d (t) −W ^T × X (t)

Ε ² = {d (t)} ² + W ^T · X (t) · X ^T (t) · W−
2d (t) ・ X ^T (t) ・ W

Equation 5] ^{E [ε 2] = E [} {d (t)} 2] + W T · E [X A (t) ·
X ^T (t)] ・ W-2E [X ^A (t) ・ d (t)] W

Therefore, in order to determine the weight vector W, the weight vector W may be determined so that the root mean square error E [ε ² ] expressed by Equation 5 is minimized. Here, X ^A (t) represents a conjugate transpose of the N-dimensional column vector X (t).

Next, the root-mean-square error E [ε expressed by equation (5)
² ] is transformed using the autocorrelation matrix Rxx represented by the following equation 6 and the correlation vector Vxxr represented by the following equation 7, the root mean square error E [ε ² ] is represented by the equation 8 Can be represented.

[0031]

Rxx = E {X ^A (t) · X ^T (t)}

Vxr = E {X ^A (t) · d (t)}

Equation 8] ^{E [ε 2] = E [} {d (t)} 2] + W T · Rxx · W-
2Vxr · W

As is apparent from Equation 8, the root-mean-square error E [ε ² ] is a concave 2 with respect to complex weights w _{1 to} w _N.
Since it is represented by the next curved surface, it surely has the minimum value, and the derivative method on the curved surface of the mean square error E [ε ² ] can be applied. Therefore, the weight vector W that makes the differential coefficient obtained by differentiating the root mean square error E [ε ² ] expressed by Expression 8 with each of the complex weights w _{1 to} w _N to 0 is zero. ² ] is minimized. Therefore, by differentiating the right-hand side of Equation 8 with each of the complex weights w _{1 to} w _N and setting its differential coefficient to 0,
The conditional expression for minimizing the root mean square error E [ε ² ] expressed by Expression 8 can be expressed by Expression 9. By modifying Equation 9, the weight vector W at that time can be expressed by the following Equation 10.

[0033]

## EQU9 ## Rxx.W-Vxr = 0

## EQU10 ## W = Rxx ^-1 · Vxr

As described above, the adaptive control processor 9 calculates the weight vector W according to the above equation 10 so as to minimize the root mean square error E [ε ² ], thereby obtaining the weight vector W. it can.

That is, the adaptive control processor 9 is configured as shown in FIG.
In the training signal period shown in FIG. 5, an autocorrelation matrix Rxx of an N-dimensional column vector X (t) having components of the input received signals x ₁ (t) to x _N (t), and a signal generator 10
Is calculated based on the reference training signal d (t) and the correlation vector Vxr between the N-dimensional column vector X (t) and the beamformer by using the above equation (10). Output to B1.

The beamformer B1 includes multipliers 2-1 to 2
-N and adder 3. In the beamformer B1, the multiplier 2-i (i = 1, 2,..., N) multiplies the received signal x _i (t) by the complex weight w _i and multiplies the signal by the adder 3. The adder 3 adds the N signals output from the multipliers 2-1 to 2-N to obtain a signal y
(T) is output. Thereby, the multipliers 2-1 to 2
The beamformer B1 including -N and the adder 3 outputs a combined signal y (t) including the desired wave and the first delayed wave. That is, when synthesizing the synthesized signal y (t), the beam former B1 simultaneously synthesizes the desired wave and the first delayed wave without separating them.

In the above description, the transmission signal transmitted from the transmitter 50, which is the desired transmitting station, has been described. However, the following description will be made including the interference wave signal transmitted from the transmitting station other than the transmitter 50. Can be explained. In the following description, all mathematical expressions used are described in the equivalent low frequency system. First, the N antenna elements 100-1 to 100-1
The received signal received by the array antenna 100 composed of 0-N includes a desired wave and a delayed wave transmitted from the transmitter 50, an inter-channel interference wave transmitted from another transmitter, and a thermal noise. It can be considered that it consists of. That is, the total number of the desired station and the interference station transmitter 50 is M, the j-th arrival wave number of L _j and for the number 2 of the received signal coming from the transmitting station N-dimensional column vector X
(T) can be expressed by the following equation (11). Where j =
It is assumed that the first transmitting station of 1 is the desired station including the transmitter 50.

[0038]

[Equation 11]

Here, H _jk is a complex variation vector of the k-th incoming wave transmitted from the j-th station, and is transmitted from the j-th transmitting station received by the i-th antenna element 100-i. It is expressed by the following _Expression 12 using the complex change amount h _ijk of the kth incoming wave. S _j (t) is the signal sequence of the arriving wave transmitted from the j-th transmitting station, and therefore, s ₁ (t) is the signal sequence transmitted from the desired station when j = 1. And s _j (t) when j ≠ 1 is a signal sequence transmitted from the interference station. Furthermore, τ _jk is j
N (t) represents a thermal noise, and a white Gaussian thermal noise _ni (t) of the i-th antenna element 100-i is a delay time of a k-th incoming wave transmitted from the i-th transmitting station. And can be represented by the following Equation 13.

[0040]

H _jk = [h _1jk , h _2jk ,..., H _Njk ] ^T

N (t) = [n ₁ (t), n ₂ (t),..., _{N N} (t)] ^T

When the weight vector W calculated during the training signal period of the desired wave is used as described above, the combined signal y (t) is expressed by the above-mentioned equation 2.
Can be represented by

That is, the signal sequence [A, B, C, D, E] included in the reproduced training signal corresponding to the desired wave and the signal sequence [A, B, C, D, E] are signal sequences [A, B, C,
D, E]. Also, the k-th delayed wave (k = 2, 3,
..) Do not match the signal sequence [A, B, C, D, E] of the training signal generated by the signal generator 10. Further, a transmission signal transmitted from a transmission station other than the transmitter 50 does not include the signal sequence [A, B, C, D, E]. Therefore,
In the receiver 30 of the communication system of the embodiment, the beamformer B1 removes the signal corresponding to the kth delayed wave (k = 2, 3, ...) And the inter-channel interference wave, and outputs the reference training signal signal. A signal corresponding to the desired wave and a signal corresponding to the first delayed wave having the same signal sequence as the sequence [A, B, C, D, E] are simultaneously output.

In the above description, it has been assumed that the first delay wave arrives later than the desired wave by one symbol period. However, in general, the first delay wave is not limited to the first delay wave, and the delay time within one symbol (L -1) (k-1) -th delayed waves (k = 2, 3,
, L) also matches the signal sequence of the reference training signal d (t) generated by the signal generator 10 of the receiver 30, so that in the communication system of the present embodiment, the receiver 30 considers all of these incoming waves as desired waves, and simultaneously synthesizes and outputs a signal corresponding to the (k-1) th delayed wave. Also in this case, a delayed wave having a delay time of one symbol or more and all inter-channel interference waves are removed. Here, if there is only one delayed wave whose delay time is within one symbol and the weight vector is ideally calculated, the composite signal y (t) output from the adder 3 is expressed by the following equation (14). Can be represented by

[0044]

[Number 14] y (t) = h1 · s 1 (t) + h2 · s 1 (t
−τ ₁₂ ) + n (t)

Here, the complex coefficients h1 and h2 are given by
The thermal noise n (t) is expressed by the following equation (17).

[0046]

[Formula 15] h1 = H ₁₁ ^T · W

[Equation 16] h2 = H ₁₂ ^T · W

[Mathematical formula-see original document] n (t) = ^NT (t) * W

The maximum likelihood estimator 7 includes a ROM (not shown) storing a maximum likelihood estimation program and a CPU (not shown) for executing the maximum likelihood estimation processing to estimate and output a digital signal included in the received signal. (Not shown) and a storage device (not shown). The maximum likelihood estimator 7 estimates a digital signal included in the received signal by executing a maximum likelihood estimation process described later in detail based on the signal y (t) output from the adder 3 of the signal controller 20. And outputs it to the demodulator 8. The demodulator 8 demodulates the digital signal output from the maximum likelihood estimator 7 and outputs a baseband signal.

The operation of the receiver 30 configured as above will be described. Antenna element 10 of array antenna 100
Each received signal received at 0-1 to 100-N is a transmission signal including a desired wave and first to (L-1) th delayed waves,
And unnecessary waves such as inter-channel interference waves. The received signals received by the antenna elements 100-1 to 100-N are received via the receiving modules RM-1 to RM-N, respectively.
The received signal R is output by the A / D converters AD-1 to AD-N.
1 to RN, in-phase distributed by the in-phase distributors 1-1 to 1-N, and input to the beam former B1.

The received signal x input to the beamformer B1
₁ (t) to x _N (t) are converted by the beamformer B1
It is formed into a composite signal including the signal received by the desired wave and the signal received by the first delay wave, and is output. Here, in the signal output from the beamformer B1, unnecessary waves such as a delayed wave arriving with a delay time of one symbol or more from a desired wave and an inter-channel interference wave are excluded. Beamformer B
The signal y (t) output from 1 is input to the maximum likelihood estimator 7.

The digital signal contained in the received signal is estimated by the maximum likelihood estimator 7 based on the signal y (t) output from the adder 3 and output, and demodulated by the demodulator 8. Then, a baseband signal is output.

Next, the maximum likelihood estimation processing in the maximum likelihood estimator 7 will be described. In the embodiment of the present invention, based on the signal y (t) output from the adder 3, the known signal Viterbi algorithm is used to estimate the maximum likelihood of the digital signal included in the received signal. Here, the Viterbi algorithm converts a received symbol sequence transmitted through a communication path that can be represented by a model using a transversal filter into an output signal of the transversal filter and a tap gain of the transversal filter. This is an algorithm for estimating the maximum likelihood based on the above. In the present embodiment, the signal y output from the adder 3
(T) is made to correspond to the output signal of the transversal filter, the tap gain of the transversal filter is estimated, and the Viterbi algorithm is applied. The number of taps of the transversal filter corresponds to the delay time of the delayed wave included in the signal y (t) output from the beam former B1.

In the following description, it is assumed that two signals, a signal received as a direct wave and a signal received as a first delayed wave, are selected by the beamformer B1. Therefore, the signal y (t) input to the maximum likelihood estimator 7 can be expressed by Expression 14.

Here, the transmitting station transmits the transmission symbol sequence {a
It is assumed that a transmission signal digitally modulated according to a digital signal composed of _n ｝ is transmitted, and that the delay time of the first delay wave is the symbol period T. Further, in the following description, the modulation method is QPSK. That is, each symbol a _n is four symbol values P (1), P (2 ), P
(3), P (4) can be taken. Here, the symbol value P
(1) = (1, 0), symbol value P (2) = (0,
1), symbol value P (3) = (-1, 0), and symbol value P (4) = (0, -1). Therefore, as shown in FIG. 10, the signal received as a direct wave and the signal received as a first delayed wave each include a received symbol sequence {Ra _n }, and the received symbols included in the delayed wave signal are: It is delayed by the symbol period T compared to the received symbol included in the desired wave signal. Therefore, the received signal Ry at the time parameter n of the signal y (t) input to the maximum likelihood estimator 7
_n can be expressed by the following equation 18. Here, the time parameter n represents a time parameter that is incremented by the symbol period T. H1 and h2 in Equation 18 are respectively Equation 1
This is a complex coefficient represented by Expression 5 and Expression 16.

[0054]

[Number 18] _{Ry n = h1 · Ra n +} h2 · Ra n-1

FIG. 3 is a flowchart of the main routine of the maximum likelihood estimation process executed by the maximum likelihood estimator 7. As shown in FIG. 3, in the maximum likelihood estimation process, the time parameter n and the undetermined symbol parameter K are initially set to 1 in step S1, and the process proceeds to step S2. In step S2, a metric and path history initial value setting process is executed, and the process proceeds to step S3. Here, in step S2, as described in detail later, the path metric PM (m) and the branch metric BM (m, 1)
And the path history PHmem (m, 1) are set to initial values. In step S3, the time parameter n is set to n +
1 is set and K is set to the undetermined symbol parameter K.
Set by substituting +1. Next, in step S4, a branch metric and path history calculation process is performed, and the flow advances to step S5. Here, in step S4, the branch metric BM (m, x) is calculated as described in detail later, and the branch metric BMmem (m, 1) and the path history PHmem (m, 1) are set. In step S5, branch metric and path history storage processing is performed, and the flow advances to step S6. Here, in step S5, as described in detail later, the path history PHmem
(I, j) and branch metric BMmem (i, j)
Is set and stored. In step S6, a path metric calculation process is executed, that is, the path metric PM (m) is calculated as described later in detail, and the flow advances to step S7.
In step S7, a received symbol determination process is performed, that is, a received symbol is determined as described later in detail, and the process proceeds to step S8. In step S8, it is determined whether or not the received signal has been completed. When the received signal has been completed, the maximum likelihood estimation process is completed. When the received signal has not been completed, the process proceeds to step S3 to proceed to step S4. S5, S
Steps S6 and S7 are repeated.

That is, in the maximum likelihood estimation processing, the processing from step S3 to step S6 is repeated for each symbol period T. In other words, every time one reception symbol Ra _n is received, the process proceeds from step S3 to step S3.
The processing up to 7 is repeated.

Next, the subroutine of the metric and path history initial value setting processing in step S2 in FIG.
This will be described in detail with reference to FIG. In this subroutine, the path metric parameter X is set by substituting the first symbol parameter of the reference symbol string stored in the ROM 11 into the path metric parameter X in step S201. Next, in step S202, the direct wave symbol parameter m is initialized to 1, and in step S203, the path history PHmem (m, 1) is set.
Is substituted for the path metric parameter X to set the path history PHmem (m, 1), and the process proceeds to step S204. In step S204, it is determined whether or not m = X.
= X, the process proceeds to step S205, while m =
If it is not X, the process proceeds to step S207. In step S205, 0 is substituted for the path metric PM (m) to set the path metric PM (m), and in step S2
06, the branch metric BMmem (m, 1) is set to 0
Is set to set the branch metric BMmem (m, 1), and the process proceeds to step S209. Step S20
In step 7, the processing maximum value is substituted into the path metric PM (m) to set the path metric PM (m), and step S2
In step 08, the processing maximum value is substituted into the branch metric BMmem (m, 1) to set the branch metric BMmem.
After setting (m, 1), the process proceeds to step S209. Here, the processing maximum value is the maximum number that can be used in the processing. Further, in step S209, m + 1 is substituted for and set to the direct wave symbol parameter m. In step S210, it is determined whether or not the direct wave symbol parameter m> 4. Returns to the main routine. On the other hand, if the direct wave symbol parameter m is not greater than 4, the process returns to step S203.

That is, in step S2, if the direct wave symbol parameter m = X, the path metric PM
(M) and the branch metric BMmem (m, 1) are both set to 0 which is the minimum value that can be taken by the path metric PM (m) and the branch metric BMmem (m, 1), and the direct wave symbol parameter m is not X In
Path metric PM (m) and branch metric BMm
em (m, 1) and path metric PM (m)
And the processing maximum value which is the maximum value that the branch metric BMmem (m, 1) can take. As described above, by initially setting the path metric PM (m) and the branch metric BMmem (m, 1) when the time parameter n = 1 to 0 or the processing maximum value, the reception symbol at the time parameter n = 2 Can be determined accurately.

Next, the subroutine of the branch metric and path history calculation processing in step S4 in FIG. 3 will be described in detail with reference to FIG. In this subroutine, in step S401, the direct wave symbol parameter m is initially set to 1, and in step S402, it is determined whether or not the undetermined symbol parameter K> 1. In this case, the process proceeds to step S403, while the undetermined symbol parameter K
If not, the process proceeds to step S404. In step S403, the symbol parameters i = 1, 2, 3, 4
For each value of the update parameter j = 1,..., K−1, the branch metric BMmem (i, j) is substituted for the branch metric BMtmp (i, j), and the branch metric BMmem (i, j) is temporarily stored. , And for each value of the symbol parameter i = 1, 2, 3, 4 and the update parameter j = 1,..., K−1, the path history PHtmp (i, j) contains the path history PHmem.
Substituting (i, j), the path history PHmem (i,
After temporarily storing j), the process proceeds to step S404.
Here, the branch metric BMtmp (i, j) and the path history PHtmp (i, j) are respectively the branch metric BMmem (i, j) and the path history PHmem (i, j) necessary for estimating the received symbol at the time n.
j) in the storage device for temporarily storing Next, in step S404, using the following equation 19, the delay wave symbol parameter x = 1,
The branch metric BM (m, x) at each value of 2, 3, and 4 is calculated.

[0060]

Equation 19] BM (m, x) = │Ry n - {h1 · P
(M) + h2 · P (x)｝ |

Here, the first term on the right side in Expression 19 is
Is the received signal Ry _n inputted to the maximum likelihood estimator 7, the second term of the right side h1 · P (m) + h2 · P (x) , the first delay and the received symbol Ra _n received by the direct wave It is assumed that the received symbols Ra _n-1 received by the waves are symbol values P (m) and P (x), respectively, and the received signal is estimated using the complex coefficients h1 and h2 of the estimated signals. This is the estimated signal at the time. Therefore, the branch metric is expressed by the number 19 BM (m, x) is the said estimated signal is the Euclidean distance between the received signal Ry _n.

Next, in step S405, step S40
4, a delay wave symbol parameter x = x1 that minimizes the sum of the branch metric BM (m, x) and the path metric PM (x) is selected, and x1 is substituted into the minimum path metric parameter Xm to minimize Set the path metric parameter Xm. That is, step S40
4, in S405, in the direct wave symbol parameter m, for all possible delayed wave symbol parameters x, that is, the received symbol Ra received in the delayed wave
_The minimum path metric parameter that calculates the branch metric BM (m, x) for all possible symbol values of _n−1 and minimizes the sum of the branch metric BM (m, x) and the path metric PM (x) Xm is determined.

Then, in step S406, the minimum path metric parameter Xm is set to the path history PHmem.
(M, 1) to substitute for the path history PHmem (m, 1).
1) is set, and in step S407, the branch metric BM is set to the branch metric BMmem (m, 1).
Branch metric BMmem by substituting (m, Xm)
(M, 1) is set. Therefore, the path history PHmem (m, 1) set in step S406 is the signal y ₁ (t) received as a direct wave at the time parameter n.
When the received symbol Ra _n is assumed to be m, the branch metric BM (m, x) and the path metric P
(X) means the received symbol Ra _n-1 of the time parameter (n-1) selected so that the value obtained by summing the sum with (x) becomes the smallest. In other words, the path history PHmem (m,
1), when the received symbol Ra _n signal y ₁ is received (t) by the direct wave at time parameter n is assumed to be m, the received symbols at time parameter (n-1) Ra _{n- It} means the symbol with the highest probability of being ₁ . The branch metric BMmem (m, 1) set in step S407 is the estimated signal at time n and the received signal Ryn when it is assumed that the received symbol R a n received by the direct wave in the time parameter _n is m. This is the Euclidean distance when the distance between is minimum.

Next, in step S408, m + 1 is substituted for the direct wave symbol parameter m and set.
In step 09, it is determined whether or not the direct wave symbol parameter m> 4. If the direct wave symbol parameter m> 4, the process returns to the main routine. Back to
Steps S404 to S408 are executed.

Next, a subroutine of the branch metric and path history storage processing in step S5 in FIG. 3 will be described with reference to FIG. In this subroutine, in step S501, it is determined whether or not the undetermined symbol parameter K> 1. If the undetermined symbol parameter K> 1, the process proceeds to step S502, while the undetermined symbol parameter K> 1. If not, the process returns to the main routine.

In step S502, symbol parameters i = 1, 2, 3, 4 and update parameters j = 2, 3,
, K, the path history PHmem (i,
j) contains the path history PHtmp @ PHmem (i,
1), the path history PHmem (i, j) is set by substituting the value stored in j−1}, and step S503 is performed.
Proceed to. Here, the path history PHmem (i, j)
Is the time parameter (n-j) selected such that the path metric PM (i) is the smallest, assuming that the received symbol R a _n received by the direct wave at the time parameter n is i. Means the symbol.

In step S503, the symbol parameters i = 1, 2, 3, 4 and the update parameters j = 2,.
For each value of K, the branch metric BMmem
(I, j) is the branch metric BMtmp @ PHm
em (i, 1), j−1}, the branch metric BMmem (i, j) is set. Then, after the process of step S503, the process returns to the main routine. Here, the branch metric BMmem
(I, j) is the sum of the branch metric BM (i, x) and the path metric P (x), assuming that the received symbol Ra _n received by the direct wave at the time parameter n is i. Means the branch metric of the time parameter (n−j + 1) selected so that the calculated value becomes smaller.

The subroutine of the path metric calculation process in step S6 in FIG. 3 will be described with reference to FIG.
In step S601, the direct wave symbol parameter m is set to 1
And the path metric PM (m) is calculated using the equation 20 in step S602.

[0069]

[Equation 20] PM (m) = BMmem (m, 1) + BMmem
(m, 2) + ... + BMmem (m, K)

[0070] Here, the path metric PM (m), when the received symbol Ra _n received by the direct wave at time n is assumed to be m, in the past the most path metric was chosen to be smaller Time n, n of symbols
-1,..., N−K + 1 branch metrics BMme
Shows the sum of m (m, 1) through BMmem (m, K). Next, in step S603, m + 1 is substituted into the direct wave symbol parameter m and set. In step S604, it is determined whether or not the direct wave symbol parameter m> 4. Return to the main routine, while the direct wave symbol parameter m
If not> 4, the process returns to step S602, and executes steps S602 and S603.

Next, the subroutine of the received symbol determination process in step S7 in FIG. 3 will be described in detail with reference to FIG. In this subroutine, step S70
In step 1, it is determined whether or not the undetermined symbol parameter K = 5. If the undetermined symbol parameter K = 5, the process proceeds to step S704. If not, the process proceeds to step S702. Proceed to.
In step S702, the path history PHmem
(1, K), PHmem (2, K), PHmem (3,
K) and PHmem (4, K) are all the same value. If they are the same value, the process proceeds to step S703, and if any one is different, the process returns to the main routine.

In step S703, the path history PH (1, K) is substituted for the received symbol Ra _{n-K + 1} and set, and then the flow advances to step S706. That is, it is determined that the received symbol Ra _{n-k + 1} is the path history PH (1, K). In step S704, symbol path parameter im that minimizes path metric PM (i) is substituted for minimum path metric parameter Y and set.
In step S705, the path history PH (Y, K) is substituted for the received symbol Ra _{n-K + 1} and set, and the process proceeds to step S706. That is, the received symbol Ra _{n-k + 1}
Is determined to be the path history PH (Y, K). In step S706, the received symbol Ra _{n-K + 1} determined in step S703 or S705 is output, and step S7 is performed.
At 07, K-1 is substituted into the undetermined symbol parameter K and set. In step S708, it is determined whether or not the undetermined symbol parameter K = 0. If the undetermined symbol parameter K = 0, the process returns to the main routine.
If the undetermined symbol parameter K is not 0, the process returns to step S702, and executes steps S702 to S707.

As described in detail above, the maximum likelihood estimator 7 including the CPU, the ROM, and the storage device performs the maximum likelihood estimation processing described above based on the signal y (t) output from the adder 3. By performing this, the reception signal is set so that the Euclidean distance between the signal estimated from the known known state transition of the digital signal included in the transmission signal and the signal y (t) output from the adder 3 is minimized. And outputs it to the demodulator 8.

In the communication system of the above embodiment, the transmitter 50 transmitting the transmission signal including the training signal shown in FIG. 3, and the desired wave and the first delayed wave arriving from the transmitter 50 correspond to the training signal 5. One symbol of each set of the set of symbol sequences is sampled so as to include a serially continuous reproduced training signal, and a reference training signal d (t (t) having a reference symbol sequence [A, B, C, D, E] ), And a receiver 30 that outputs a composite wave of the desired wave and the first delayed wave based on the reproduced training signal and the reference training signal d (t). A transmission signal from a desired transmitter 50 can be received.

<Modification> In the above embodiment, as shown in FIG. 3, the transmitter 50 includes the first to fifth symbols each including two identical symbols A, B, C, D, and E, respectively. A transmission signal including a training signal in which the symbol sequence is serially continuous is transmitted. However, the present invention is not limited to this, and as shown in FIG.
A configuration is also possible in which a transmission signal including a training signal in which the first to fifth symbol strings each including three identical symbols A, B, C, D, and E are serially continuous is transmitted. Good. In this case, as shown in FIG. 19, the adaptive control processor 9 sets the time of the third symbol in the first symbol sequence of the training signal included in the transmission signal of the desired wave received first to the reference time. The reproduction training signal is reproduced by sampling at the reference time and the sampling period Ts equal to the time length of the symbol sequence from the reference time. Even with the above configuration, the same effects as those of the embodiment can be obtained, and a delayed wave arriving with a delay time within two symbol times from the desired wave can be received. In this modification, the time of the third symbol in the first symbol sequence is used as a reference, but the time of the second symbol may be used as a reference.

In the above embodiment, the transmitter 50
As shown in FIG. 3, two identical symbols A,
A transmission signal including a training signal in which the first to fifth symbol sequences including B, C, D, and E are serially continuous is transmitted. However, the present invention is not limited to this, and the first to fifth symbol trains each including four or more P plural identical symbols A, B, C, D, and E are serially continuous. You may comprise so that the transmission signal containing a signal may be transmitted. In this case, the adaptive control processor 9 sets the reference time and the reference time as the reference time, preferably the time of the P-th symbol in the first symbol sequence of the training signal included in the transmission signal of the desired wave received first. The reproduction training signal is reproduced by sampling at a sampling period Ts equal to the time length of the symbol sequence from the reference time. Even with the above configuration, the same effects as those of the embodiment can be obtained, and a delayed wave arriving with a delay time within (P-1) symbol time from the desired wave can be received. In this modification, the time of the P-th symbol in the first symbol sequence is used as a reference, but the time of at least the second and subsequent symbols may be used as a reference.

In the above embodiment, a training signal composed of five serially-consecutive symbol strings is used. However, the present invention is not limited to this, and a training signal composed of at least two or more symbol strings is used. You may. Even with the above configuration, the same effects as in the embodiment can be obtained.

In the above embodiment, a training signal is used in which the first to fifth symbol sequences each consisting of two identical symbols A, B, C, D and E are serially continuous. Although the present invention is not limited to this, the present invention is not limited to this.
The fifth to fifth symbol sequences may include a signal that is serially continuous. Even with the above configuration, the same effects as in the embodiment can be obtained.

In the above embodiment, the symbols A, B, C, D, and E in each symbol row are set differently. However, the present invention is not limited to this, and may be set to any symbol row. Further, for example, the symbols may be set so as to be different at least between adjacent symbol columns.
Even with the above configuration, the same effects as those of the embodiment can be obtained.

In the above embodiment, the adaptive control processor 9 determines whether the autocorrelation matrix R of the N-dimensional column vector X (t) is
xx and a so-called SMI (S) that calculates a weight vector W based on a correlation training vector Vxr between the reference training signal d (t) and the N-dimensional column vector X (t).
The calculation is performed using a calculation method called an "amplified matrix inverse method". However, the present invention is not limited to this, and the recursive least squares RLS (Recursive Least S
LMS (Least Mean Square) algorithm, which is a least-squares algorithm or a least mean square algorithm.
The adaptive control processor 9 may be configured to calculate the weight vector W using another algorithm such as an algorithm. Further, the LMS algorithm may be any of a gradient method, a re-steep descent method, and a Neaton method. Even with the configuration described above, the same operation as in the embodiment is performed and the same effect is obtained.

[0081]

DESCRIPTION OF THE PREFERRED EMBODIMENTS In order to confirm the operation of the receiver 30 in the communication system of the embodiment, the present inventors performed a simulation described below and calculated the average bit error rate of the receiver 30. For comparison, a conventional adaptive array antenna and a prior art document 2 "Yoshiharu Doi, Takeo Ohgane, Eiichi Ogawa," ISI based on a combination of an adaptive array antenna and a maximum likelihood equalizer in a quasi-static Rayleigh fading channel.・ Characteristics of CCI adaptive canceller, “IEICE Technical Report, vol. RCS95-46, p.
p. 19-224, July 1995 ”, a computer simulation was also performed for the system proposed. In the method of the related art document 2, the MMSE-type adaptive array antenna uses a plurality of training signals corresponding to the delay time of each arriving wave from a desired station, and creates an antenna pattern corresponding to each arriving wave. This is a method in which once separated signals are recombined, and then the maximum likelihood estimation is performed. FIG.
4 to 16, the result of the simulation by the method of the related art document 2 is shown as a comparative example.

In this simulation, a propagation path model as shown in FIG. 12 was employed. In FIG. 12, the horizontal axis represents the relative time standardized by the transmission symbol, and the vertical axis represents the average power of the arriving wave. The oblique axis indicates the station that simply transmitted each incoming wave. Receiver 3
0 is assumed to receive multipath delayed waves from one desired station and (M-1) interfering stations, and each incoming wave is assumed to be subjected to independent Rayleigh fading for each path. Since a quasi-static Rayleigh fading propagation path is assumed here, the propagation path does not fluctuate between bursts, and a model is assumed in which the next burst receives fading fluctuations completely independent of the previous burst. doing. It is assumed that the first signal arriving from each interfering station is delayed from the arrival time of the desired wave. The offset time of the arrival time between the desired wave and the interference wave will be described later.

In this embodiment, the TDMA / QPS
The communication system of the embodiment was applied to the K system, and computer simulation was performed. In the computer simulation, as shown in FIG. 13, the time slot configuration was such that the time slot length was 128 symbols, of which the training signal sequence was 16 symbols and the data signal sequence was 112 symbols. It is assumed that there is no correlation of fading fluctuation between each antenna element. Training signals used by the desired station and the interfering station selected signals having good orthogonality to each other. The weight vector is calculated by using SMI (S
Calculated based on the method of [Amplified Matrix Inverse]. The maximum likelihood estimation algorithm includes the Viterbi algorithm (Viterbi Al
gorism) was employed. Since the signal y (t) input to the maximum likelihood estimator 7 is two waves, the number of Viterbi states is set to four.

The number N of antenna elements was set to be the same as the number of incoming waves, and the degree of freedom of the adaptive array was set to zero. In this specification, C / I is indicated by the ratio of the total power of the arriving wave from the desired station to the total power of the arriving wave from the interference station.

There is one desired station and one interfering station. In the propagation path where two waves (first arriving wave + 1 symbol delayed wave) multipath propagate from each station, the desired wave and the inter-channel interference wave Time of arrival offset between 1
When the symbols are used, the average C / I is 0 dB and 5 respectively.
Average signal-to-noise power ratio E for dB and 10 dB
FIGS. 14 and 1 show b / N ₀ versus the average bit error rate, respectively.
5, shown in FIG.

As is clear from FIGS. 14 to 16, the receiver 30 operates well and the average bit error rate due to the interference wave is obtained when the average C / I is 0 dB, 5 dB, or 10 dB, respectively. You can see that you can reduce. That is, the communication system according to the present invention shows superior characteristics in all cases as compared with the conventional example and the comparative example. 14 to 16, the average signal-to-noise power ratio Eb / N ₀ is 10d.
Above B, the slope of the graph {(average bit error rate) / (average signal-to-noise ratio Eb / N ₀ )} is (one digit / 5 dB), which results in a path diversity gain of two branches. You can see that it is. Also in comparison with the comparative example, the difference appears in the slope of the graph when the average C / I is low as shown in FIG.

This can be explained as follows. A complex variation vector H ₁₁ of the desired wave, if accidentally correlated to one-symbol delay delayed wave complex variation vector H ₁₂ in occurs, the system of the comparative example to separate once all the paths,
When selecting one of the paths, the directivity null is directed to the other path having a strong correlation, so that the S / N of the synthesized signal is deteriorated. On the other hand, in the present system, the desired wave and 1
Since the symbol delayed waves are simultaneously captured, the S / N of the composite signal y (t) does not deteriorate even in such a case. Therefore, it is considered that the present system can obtain good characteristics even at a low C / I.

Next, the arrival time offset characteristic will be described. As described above, the characteristics of the communication system according to the present embodiment vary depending on the arrival time difference between the desired wave and the first arrival wave from the interference station. If the arrival time difference is 0, the inter-channel interference wave arriving first and the inter-channel interference wave delayed by one symbol have exactly the same signal components in the training signal period. At this time, the receiver 30 may operate so that the combined component of the first inter-channel interference wave and the one-symbol-delayed inter-channel interference wave in the training period becomes zero. You may not be able to turn your gender null. When the receiver 30 operates in this way, these inter-channel interference waves cannot be canceled each other during the data signal period, so they remain as interference components, and the average bit error rate deteriorates.

Therefore, in order to examine the influence of the arrival time offset, the average signal-to-noise ratio Eb / N ₀ is 10 dB, 1
In each of the cases of 5 dB and 20 dB, the average bit error rate characteristics with respect to the arrival time difference were evaluated. The result is shown in FIG. Arrival time differences on the horizontal axis are represented by times standardized by the symbol length. Circles represent average signal to noise ratio Eb /
N ₀ = 20 dB, triangles indicate average signal-to-noise ratio Eb / N ₀ = 1
5 dB, squares indicate average signal-to-noise ratio Eb / N ₀ = 10 dB
It is the characteristic at the time of. From FIG. 17, it can be seen that the characteristics are worse at even time points. This result can be explained as follows.

When the arrival time difference is an even timing other than 0, a part of the training signal of the first arriving inter-channel interference wave coincides with the training signal of the one-symbol-delayed inter-channel interference wave. At this time, if the remaining signals in the training period coincide with each other by chance, the characteristics deteriorate as in the operation when the arrival time offset is 0. As time elapses, the probability that all the signals in the training period match becomes smaller, so that the range of the characteristic deterioration becomes smaller. From FIG. 17, when the average signal-to-noise ratio Eb / N ₀ = 20 dB and the arrival time difference is 0.3 symbol or more, the average bit error rate is about 3 × 10 ⁻³ at the worst value, and 10 on average. ^-4 , sufficient characteristics can be obtained if the arrival time difference is 0.3 symbols or more.

As described above, the receiver 30 in the communication system of the present embodiment requires an arrival time offset of 0.3 symbol or more between the desired wave and the interference wave. However, the time of 0.3 symbol is, for example, 2 Mbp
s TDMA / QPSK cellular system, assuming 0.3 μs. Therefore, even if the cellular system synchronizes between cells, the present communication system indicates that the same frequency can be reused in cells separated by about one cell.

From the results of the above simulation, the effects of the present embodiment can be summarized as follows. (1) An interference wave can be satisfactorily removed. (2) A sufficient path diversity gain can be obtained even when the average C / I is low. (3) It can be applied to a cellular system that synchronizes between cells.

[0093]

According to the first aspect of the present invention, there is provided a radio communication system including the first training signal, wherein a plurality of Q sets of symbol sequences each including a plurality of P identical symbols are serially continuous. A wireless transmitter having a transmitting means for transmitting a signal; a second training signal in which the first training signal and one symbol of each of the plurality of Q symbol sequences are serially continuous. And a wireless receiver having a receiving means for receiving based on the training signal, wherein the wireless receiver is configured to include a received wireless signal in a wireless signal of a direct wave. The time after the second symbol in the first symbol sequence of the first training signal is set as a reference time, sampling is performed at a sampling interval equal to the time length of each symbol sequence. Sampling means for outputting a third training signal comprising a sequence of coupled symbols; and, based on a correlation between the second training signal and the third training signal, converting the received wireless signal into Signal control means for combining at least a direct wave and a first delayed wave signal which arrives first after being delayed from the direct wave signal and outputs the combined signal as a received signal.
Thus, the radio receiver in the radio communication system can simultaneously combine the direct wave signal and the first delayed wave without deteriorating the signal-to-noise ratio S / N of the combined signal, and appropriately remove unnecessary waves. And the error rate in the receiver can be reduced as compared with the conventional example.

According to a second aspect of the present invention, there is provided the radio communication system according to the first aspect, wherein the reference time is equal to a first symbol sequence of a first training signal included in the direct wave radio signal. The time is set to the time of the P-th symbol. Thus, the wireless receiver in the wireless communication system according to the second aspect can combine a direct wave and a delayed wave arriving from the direct wave with a delay time shorter than the time length of the symbol sequence and output the combined signal as a received signal. .

Further, the wireless communication system according to claim 3 is the wireless communication system according to claim 1 or 2,
The receiver includes an array antenna composed of a plurality of N antenna elements that are closely arranged in a predetermined arrangement shape, and the signal control unit controls the third training signal and the second training signal. And an adaptive control means for calculating and outputting the N-dimensional weight vector based on the correlation vector of the third training signal and the autocorrelation matrix of the third training signal; Beam forming means for simultaneously synthesizing the direct wave signal and the at least one delayed wave signal by multiplying an N-dimensional column vector having the received signal as a component by the weight vector, and outputting as a received signal And Thus, the wireless receiver in the wireless communication system according to claim 3 removes the delayed wave and the interference wave that arrive from the direct wave with a delay equal to or more than the time length of the symbol sequence, and removes the direct wave and the direct wave. At least the first delayed wave signal that arrives first after being delayed from the wave signal can be combined and output as a received signal.

According to a fourth aspect of the present invention, there is provided the radio receiver according to the fourth aspect, wherein the plurality of Q sets of symbol sequences each including the plurality of P identical symbols include a first training signal which is serially continuous. Receiving means based on the first training signal and a second training signal in which one symbol of each of the plurality of Q symbol sequences is serially continuous. Wherein the radio signal received by the receiving means is defined as a time after a second symbol in a first symbol sequence of a first training signal included in a direct wave radio signal as a reference time. A third training signal including a reference time and a sequence of symbols sampled at a sampling interval equal to the time length of each symbol sequence from the reference time is output. Sampling means for delaying the radio signal received by the receiving means based on the correlation between the second training signal and the third training signal, And a signal control unit that combines at least the first delayed wave signal that reaches the first signal and outputs the combined signal as a received signal. Thus, the radio receiver can simultaneously combine the direct wave signal and each of the delayed wave signals, appropriately remove unnecessary waves, and reduce the error rate as compared with the conventional example.

According to a fifth aspect of the present invention, in the wireless receiver according to the fourth aspect, the reference time is equal to the first time of the first training signal included in the first received direct wave radio signal. Is set at the time of the P-th symbol in the symbol sequence. Thus, the wireless receiver according to claim 5 can combine a direct wave and a delayed wave arriving from the direct wave with a delay time shorter than the time length of the symbol sequence and output the combined signal as a received signal.

Further, the radio receiver according to claim 6 is the radio receiver according to claim 4 or 5, further comprising an array antenna comprising a plurality of N antenna elements closely arranged in a predetermined arrangement shape. Wherein the signal control means comprises:
The beam forming is performed by calculating the N-dimensional weight vector based on a correlation vector between the third training signal and the second training signal and an autocorrelation matrix of the third training signal. Means for outputting to the means, and an N-dimensional column vector having N received signals received by each of the antenna elements as a component, multiplied by the weight vector to obtain the direct wave signal and at least the direct wave signal. Beam forming means for synthesizing one delayed wave signal and outputting it as a reception signal. Accordingly, the wireless receiver according to claim 6 removes the delayed wave and the interference wave that arrive from the direct wave with a delay longer than the time length of the symbol sequence, and delays the direct wave and the direct wave signal. Then, at least the first delayed wave signal that arrives first can be combined and output as a received signal.

[Brief description of drawings]

FIG. 1 is a block diagram of a receiver 30 in a communication system according to an embodiment of the present invention.

FIG. 2 is a block diagram of a transmitter 50 in the communication system according to the embodiment of the present invention.

FIG. 3 is a diagram showing a configuration of a time slot of a transmission signal in the embodiment according to the present invention.

FIG. 4 is a diagram showing a sampling period Ts and a reproduction training signal of a desired wave and a delayed wave in the embodiment according to the present invention.

FIG. 5 is a flowchart of a main routine of a maximum likelihood estimation process executed by the maximum likelihood estimator 7 of FIG. 1;

FIG. 6 is a flowchart of a subroutine of a metric and path history initial value setting process of FIG. 5;

FIG. 7 is a flowchart of a subroutine of a branch metric and path history calculation process of FIG. 5;

FIG. 8 is a flowchart of a subroutine of a branch metric and path history storage process of FIG. 5;

FIG. 9 is a flowchart of a subroutine of a path metric calculation process in FIG. 5;

10 is a flowchart of a subroutine of received symbol determination processing of FIG.

11 is a diagram illustrating a reception symbol sequence included in a signal y ₁ (t) received by a desired wave in the receiver 30 of FIG.
FIG. 9 is a diagram illustrating a received symbol sequence included in a signal y ₂ (t) received as a delayed wave.

FIG. 12 is a diagram illustrating a propagation path model in the embodiment.

FIG. 13 is a diagram illustrating a configuration of a time slot of a transmission signal used in the embodiment.

14 is a graph showing an average bit error rate with respect to an average signal-to-noise power ratio when the average C / I is 0 dB in the receiver 30 of the embodiment in FIG. 1;

FIG. 15 is a graph showing an average bit error rate with respect to an average signal-to-noise power ratio when the average C / I = 5 dB in the receiver 30 of the embodiment of FIG. 1;

FIG. 16 is a graph showing an average bit error rate with respect to an average signal-to-noise power ratio when the average C / I = 10 dB in the receiver 30 of the embodiment of FIG. 1;

17 is a graph showing an average bit error rate with respect to an arrival time difference between a desired wave and a first delayed wave in the receiver 30 of the embodiment of FIG.

FIG. 18 is a diagram showing a configuration of a time slot of a transmission signal in a modification.

FIG. 19 is a diagram showing a sampling period Ts and a reproduction training signal of a desired wave and a delayed wave in a modified example.

[Explanation of symbols]

 1-1 to 1-N: In-phase distributor; AD-1 to AD-N: A / D converter; RM-1 to RM-N: receiving module; B1: beam former; 2-1 to 2-N ... Multiplier, 3 ... Adder, 7 ... Maximum likelihood estimator, 8 ... Demodulator, 9 ... Adaptive control processor, 10 ... Signal generator, 11 ... ROM, 20 ... Signal controller, 30 ... Receiver, 50 ... Transmitter, 100: array antenna, 100-1 to 100-N: antenna element.

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification number Reference number within the agency FI Technical indication location H04L 27/18 H04L 27/00 B (72) Inventor Eiichi Ogawa Kyotani, Seika-cho, Soraku-gun, Kyoto Prefecture Mihiratani No. 5 in the sub-section of ATR Optical Radio Communications Laboratory

Claims (6)

[Claims] 1. A wireless transmitter having a transmitting means for transmitting a wireless signal including a predetermined first training signal and a subsequent data signal, and a wireless signal transmitted from the wireless transmitter, A wireless communication system comprising: a wireless receiver including receiving means for receiving based on a first training signal included in a wireless signal and a predetermined second training signal; The training signal includes a plurality of Q sets of symbol sequences each including a plurality of P identical symbols serially and continuously, and the second training signal includes one symbol of each set of the plurality of Q sets of symbol sequences. Are serially continuous, and the wireless receiver converts the wireless signal received by the receiving means into a first signal included in the first received direct wave wireless signal. The time after at least the second symbol predetermined in the first symbol sequence of the training signal as a reference time is sampled at the reference time and at a sampling interval equal to the time length of each symbol sequence from the reference time. Sampling means for outputting a third training signal comprising a sequence of symbols sampled and sampled, and a signal received by the receiving means based on a correlation between the second training signal and the third training signal. Signal control means for synthesizing at least the direct wave and the first delayed wave signal which arrives first after delaying the direct wave signal from the direct wave signal, and outputting the combined signal as a received signal. Wireless communication system. 2. The method according to claim 1, wherein the reference time is a time of a P-th symbol in a first symbol sequence of a first training signal included in a radio signal of a direct wave received first. The wireless communication system according to claim 1. 3. The radio receiver according to claim 1, further comprising an array antenna including a plurality of N antenna elements juxtaposed and arranged in a predetermined arrangement shape, wherein the signal control means includes: the third training signal; Based on a correlation vector between the second training signal and the autocorrelation matrix of the third training signal, so as to synthesize only a signal in which the third training signal matches the second training signal. Adaptive control means for calculating and outputting an N-dimensional weight vector for weighting; and a weight vector for the N-dimensional column vector having N received signals received by each of the antenna elements as components. A beam that combines a signal including the direct wave signal and the at least one delayed wave signal and outputs the combined signal as a received signal The wireless communication system according to claim 1 or 2, characterized in that a forming means. 4. The method according to claim 1, further comprising: transmitting a radio signal transmitted from a radio transmitter transmitting a radio signal including a predetermined first training signal and a subsequent data signal to a first training signal included in the radio signal. The second predetermined
And a receiving means for receiving the first training signal based on the first training signal, wherein the first training signal includes a plurality of Q sets of symbol sequences each including a plurality of P identical symbols, which are serially continuous. In the second training signal, one symbol of each set of the plurality of Q sets of symbol sequences is serially continuous, and the wireless receiver converts the wireless signal received by the receiving unit into: The time after at least a predetermined second symbol in the first symbol sequence of the first training signal included in the first received direct wave radio signal is set as a reference time, the reference time and the reference A third trainee consisting of a sequence of sampled symbols sampled at a sampling interval equal to the time length of each symbol sequence from the time Sampling means for outputting a radio signal received by the receiving means on the basis of a correlation between the second training signal and the third training signal. And a signal control means for combining at least a first delayed wave signal that arrives first with a delay and outputs the received signal as a received signal. 5. The method according to claim 5, wherein the reference time is a time of a P-th symbol in a first symbol sequence of a first training signal included in a radio signal of a direct wave received first. The wireless receiver according to claim 4. 6. The radio receiver further includes an array antenna composed of a plurality of N antenna elements juxtaposed and arranged in a predetermined arrangement shape, wherein the signal control means includes: the third training signal; Based on a correlation vector between the second training signal and an autocorrelation matrix of the third training signal, only a signal in which the third training signal matches the second training signal is synthesized. An adaptive control means for calculating and outputting an N-dimensional weight vector for weighting the weights, and an N-dimensional column vector having N received signals received by each of the antenna elements as components. By multiplying a vector, a signal including the direct wave signal and the at least one delayed wave signal is synthesized and output as a received signal. Radio receiver of claim 4 or 5, wherein further comprising a chromatography beam forming means.