JP2591241B2 - Adaptive diversity receiver - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) The present invention is intended to efficiently remove intersymbol interference and improve data transmission characteristics when performing data transmission via a communication path in which time-varying intersymbol interference occurs. The present invention relates to an adaptive diversity receiving apparatus to be improved.

(Prior Art) In a communication system in which time-varying intersymbol interference occurs, there is a method of improving reception characteristics by an adaptive reception method combining a diversity method and an adaptive equalization method. As a conventional example of such a technique, a decision feedback equalization method is used as an equalization method, a matched filter or a forward equalizer is provided in each diversity branch, and the outputs thereof are combined and input to a decision feedback equalizer ( For example, Watanabe,
"Adaptive Receiving Method in Multipath Channel", IEICE Technical Report, CS78-203, pp.57-64, p.Monsen, "Adaptive Equalization of the Slow Fading Channel", IEE, Transaction of Communications, Vol.COM-22, No.
8, Aug. 1974) is known.

Also, as an optimal equalization method, maximum likelihood sequence estimation (Maximu
An equalization method using the m Likelihood Sequence Estimation (MASE) method is known (for example, by Proakis,
“Digital Communications”, McGraw Hill, 19
83). One well-known method of implementing this method is to use a Viterbi algorithm (for example, Hayes, “The Viterbi Algorithm Applied Digital Data Transmission, IEE”).
EE, Communication Society, 1975, N
o.13, pp.15-20). In this method, a constant uniquely determined by the output of the matched filter and the communication path (branch metric constant part: Haze, “The Viterbi Algorithm Applied Toe Digital Data Transmission, IEE,
Society, 1975, No.13, pp.18, equation 8b), right-hand side second
Term and the third term), a branch metric of the received signal at each reception time point is obtained. As a conventional technique combining such an equalization scheme and diversity reception, a scheme is proposed in which the quality of a received signal (spread of intersymbol interference, SN ratio, etc.) in each diversity branch is estimated, and a selection is made based on the estimated quality. (For example, Okanoue and Furuya, "New detection selection diversity suitable for MLSE", IEICE Autumn National Convention 1989, B
-502). Further, a reception method (adaptive MLSE) which is extended to perform an adaptive operation so as to cope with time-varying intersymbol interference is also known (for example, Proakis, Digital Communications ", McGraw H
ill, 1983). As an adaptive reception system combining this reception system with the diversity system, one having the configuration shown in FIG. 6 can be considered. In FIG. 6, the signals received from the diversity branch 600 are combined, and the combined signal is input to the above-mentioned adaptive MLSE receiver.

(Problems to be Solved by the Invention) However, when the configuration as shown in FIG. 6 is used, a signal from a branch having a low received signal level and a bad SN ratio and a signal from a branch having a high received signal level and a good SN ratio are used. Are evaluated with the same weight. For this reason, in a mobile communication system or the like in which deep fading occurs such that the received signal level hardly disappears, an error occurs in the adaptive operation of the MLSE, and the reception characteristics of the entire diversity receiving system deteriorate.

(1) An adaptive diversity receiving apparatus using a plurality of (L) antennas according to the present invention is connected to each diversity branch, and receives each received signal in each diversity branch. And after-mentioned judgment results are input, the channel impulse response to each of the diversity branches is estimated, and the estimation result of the channel impulse response to each of the diversity branches and the internal state used to obtain the estimation result are output. L communication channel impulse response estimating circuit groups to be connected to each of the diversity branches, receiving the received signal in each of the diversity branches and the estimation result of the communication channel impulse response up to each of the diversity branches, and inputting each of the diversity branches Smell L branch metric operation circuits for obtaining a branch metric of a received signal to be received, and outputs from the L communication channel impulse response estimating circuits as input signals, and branching the received signals in each of the diversity branches based on the input signals. A branch metric quality operation circuit for estimating and outputting the metric quality, an output of the L branch metric operation circuits and an output of the branch metric quality operation circuit, and a branch metric for a received signal in each of the diversity branches A branch metric combining circuit that combines a branch metric for the received signal in each of the diversity branches based on the quality of the diversity branch and outputs the result as a branch metric of the entire diversity; As input switch metric performs maximum likelihood sequence estimation based on the diversity overall branch metric, and a maximum likelihood sequence estimation circuit for outputting a determination result.

(2) In the adaptive diversity receiver according to item (1), each of the L communication channel impulse response estimating circuits receives the determination result described in item (1) as input, and outputs an input signal to each tap. Output to the adaptive control processor described below, and each tap coefficient _k (i) (k
= 1, 2,... L) and a transversal filter composed of M taps for input and output to and from the adaptive control processor, and a received signal in each of the diversity branches,
A delay circuit for delaying an input signal by a demodulation time generated in the maximum likelihood sequence estimation circuit, an output of the delay circuit and an output of the transversal filter as input signals, and an error signal of the input signal for adaptive control described later. A processor, a subtraction circuit that also outputs to an output terminal as an internal state used for obtaining the estimation result, and the subtraction circuit output,
An input signal to each of the taps and each of the tap coefficients are input, and an estimation result _k (i + 1) _obtained by estimating a channel impulse response to each of the diversity branches at time i + 1 is set as an output terminal and each of the tap coefficients. An adaptive control processor for outputting to the transversal filter.

(3) In the adaptive diversity receiver according to (2), the branch metric quality calculation circuit inputs L error signals output from the communication path impulse response estimation circuit connected to each diversity branch. When the L signal power detection circuit groups for obtaining the signal power of the input signal and the respective outputs of the L signal power detection circuit groups are input, and the input signal is smaller than a predetermined level, It outputs a normal signal, outputs an alarm signal when it is large, and is controlled by a control signal from a comparator control signal described later so as to continuously output the alarm signal when the alarm signal is output once. L comparators and a signal output from the L comparators are input to the comparator which once outputs the alarm signal. And a comparator control circuit for outputting the control signal Te.

(4) In the adaptive diversity receiver according to (2), the branch metric quality operation circuit estimates a channel impulse response to each of the diversity branches connected to each of the diversity branches, and the estimation result _k (i). (K = 1, 2,... L), and L delay circuit groups for delaying and outputting respective input signals by one time, and respective outputs of the L delay circuit groups
_k (i) and an estimation result _k (i + 1) _obtained by estimating the channel impulse response up to each of the diversity branches at time i + 1 are input, and an input signal vector
The difference vector Δ _k between _k (i) and _k (i + 1) =
_k (i + 1) _-k (i) is obtained, and the difference vector Δ
The absolute value of each element of _k is determined, and among the absolute values of each element,
The L channel impulse response variation calculation circuit group that outputs the largest one and the outputs of the L communication channel impulse response variation calculation circuit groups are input, and the input signal is higher than a predetermined level. If it is small, it outputs a normal signal, if it is large, it outputs an alarm signal, and if it outputs the alarm signal once according to the control signal from the threatening device control signal, it keeps outputting the alarm signal. Comparators controlled as described above, and a comparator control that inputs signals output from the L comparators and outputs the control signal to the comparator that once output the alarm signal And a circuit.

(5) In the adaptive diversity receiver according to (3) or (4), in the branch metric synthesis circuit, each output signal from the L comparators is input, and the input signal is equal to the normal signal. Has 1,
L signal conversion circuit groups that output 0 when they are equal to the alarm signal, and (b) a first addition circuit that inputs and adds the respective outputs of the L signal conversion circuit groups,
2. A group of L multipliers for multiplying each output of the group of L signal converters and each of the branch metrics obtained from each diversity branch according to claim 1 connected to the group of signal converters. A second addition circuit for adding the outputs of the L multiplication circuit groups, an output of the addition circuit 1 and an output of the second addition circuit, and an output of the second addition circuit. A dividing circuit for dividing by an output of the first adding circuit.

(Operation) In a diversity branch in which the SN ratio is degraded or an error occurs in the estimation of the channel impulse response, a large error occurs in estimating the channel impulse response. By using the estimation process information at the time of the impulse response estimation, it is possible to estimate a branch whose received signal quality is degraded. According to the present invention, branch metrics obtained from each diversity branch are synthesized based on the communication path impulse response estimation process information in each diversity branch. For this reason, it is possible to suppress the influence that occurs in the branch where the received signal quality is degraded, and to improve the receiving characteristics of the entire receiving system.

(Embodiment) FIG. 1 is a system diagram showing the principle of an adaptive diversity reception system according to the first invention of the present application. In the figure,
1000 is an L diversity branch group, 1001 is a channel impulse response estimation circuit group, 1002 is a branch metric operation circuit group, 1003 is a branch metric quality estimation circuit, 1004 is a branch metric synthesis circuit, 1005 is a soft decision Viterbi demodulation circuit, 106 is an output terminal.

The received signals received by the diversity branch group 1000 are respectively the channel impulse response estimation circuit group.
1001 and the branch metric calculation circuit group.
Each circuit of the communication channel impulse response estimation circuit group 1001 can be configured using, for example, an M-tap transversal filter as shown in FIG. 2 (for example, “Digital Communications” by Proakis, McGraw Hill , 1983).

FIG. 2 shows an embodiment of a communication channel impulse response estimation circuit. In FIG. 2, the determination result is sequentially input from an input terminal 200 to a register 203 and a multiplier 204, and
1 receives a received signal. The adder 205 has a multiplier 204
Is output, and a received signal replica configured based on the determination result and the estimated value of the channel impulse response is output from the output. At this time, the timing of the received signal replica and the actual received signal can be matched by delaying the received signal input from the input terminal 201 by the demodulation delay in the delay circuit 208 (for example, “Proakis, Digital Communications ”,
mcGraw Hill, 1983). The estimated value _k (i + 1) of the channel impulse response at time i + 1 (k = 1, 2,...,
L) is the estimated value _k (i) of the channel impulse response at time I one time before, and the input signal to each tap.
based on the error signal e _k (i) in the k-th diversity branch obtained from _{k (i)} and the subtraction circuit 206, obtained by the processor 207. Where _k (i)
Is an M-dimensional vector in which each tap coefficient (202) of the k-th diversity branch at time i is an element, and _k (i) is an input signal (judgment result) to each tap at time i as an element. This is the obtained M-dimensional vector.
The processor 207 uses, for example, the algorithm shown in Equation (6.7.56) of “Digital Communications” written by Proakis, McGraw Hill, 1983, to obtain _k (i
+1) is obtained and output from the output terminal 210. Furthermore, e _k
(I) is output from the output terminal 211 as the internal state used to obtain the estimation result.

In each circuit of the branch metric calculation circuit group 1002 in FIG. 1, each input signal from the diversity branch group 1000 and the current communication channel impulse response estimation value output from each of the communication channel impulse response estimation circuit group 1001 _{Based on k} (k = 1, 2,..., L), a branch metric of the received signal in each diversity branch is obtained. Here, the branch metric of the received signal in each diversity branch is, for example, Haze, “The Viterbi Algorithm Applied Toe Digital Data Transmission” IEEE, Communication Society, 1975, No.13, pp.18. 8b) Determined as shown on the right side of the equation.

On the other hand, the output signal of the channel impulse response estimation circuit group 1001 is also input to the branch metric quality estimation circuit 1003. The branch metric quality estimation circuit 1003
For example, it can be configured as shown in FIG. 3 and FIG.

In the configuration of FIG. 3, error signals output from each circuit of the communication channel impulse response estimation circuit 1001 connected to each diversity branch 1000 are input from the input terminal group 300, and the power of each error signal is detected by the power detection circuit group. Detect at 302. Here, in the diversity branch where the detected error signal power is large, it is considered that the channel impulse response is not properly estimated. Here, for example, a power level such that the level of the signal giving the detected error signal power exceeds the determination region is set in advance as a threshold level, and each output of the power detection circuit group 302 and the threshold level are compared with the comparator group 303. To compare. In the comparator group 303, when the input signal is smaller than the threshold level, the channel impulse response is normally estimated, and it is determined that the quality of the branch metric obtained based on the estimated channel impulse response is good. Outputs a normal signal. Conversely, if the input signal is larger than the threshold level, it is determined that the quality of the branch metric is poor, and an alarm signal is output. The output of the comparator group 303 obtained in this way is
Output to output terminal group 9305 and comparator control circuit
Also output to 304. The comparison circuit control circuit 304 controls the comparator that has output the alarm signal once so as to always output the alarm signal. This control is performed to remove the influence from the quality of the branch metric because the channel impulse response estimating circuit 1001 may diverge once the channel impulse response estimation fails. As described above,
The quality of the branch metric of each diversity branch is output from the output terminal group 305.

In the configuration of FIG. 4, an estimated value of the channel impulse response output from the channel impulse response estimation circuit group 1001 is input from the input terminal group 400. The estimated value of the channel impulse response from each channel impulse response estimation circuit is input to the delay circuit group 401. The delay circuit group 401 delays the input signal by one time (reference time) and outputs the delayed signal to the communication path impulse response variation calculation circuit group 402, respectively. Further, the estimated value of the current (time i) communication channel impulse response _k (i) input from the input terminal group 400 is also input to the communication channel impulse response variation calculation circuit group 402. As described above, each communication channel impulse response variation calculation circuit group 402 includes the current communication channel impulse response estimation value _k (i) in each diversity branch and the current communication channel impulse response estimation in each diversity branch one time before. The value _k (i-1) is input. In each circuit of the channel impulse response variation calculating circuit 402 calculates a difference vector delta _k of the input vector _{k (i), k (i} -1). Further, the absolute value of each element of the delta _k, and outputs the maximum value. By the above operation, the maximum level of the fluctuation generated within the reference time of the communication path impulse response in each diversity branch is output from each circuit of the communication path impulse response fluctuation calculation circuit group 402. Here, assuming that the reference time is one symbol time, the fluctuation speed of the communication channel impulse response in each diversity branch is usually sufficiently smaller than the baud rate. Therefore, when the output of each circuit of the communication channel impulse response variation calculation circuit group 402 is larger than the level difference between the modulation signal points, it is considered that the estimation of the communication channel impulse response was not successful. Here, for example, the minimum level difference between the modulation signal points is set in advance as a threshold level, and the output of each circuit of the communication path impulse response fluctuation estimation circuit group 402 is compared with the threshold level by the comparator group 403. In the comparator group 403,
If the input signal is smaller than the threshold level, the channel impulse response is normally estimated, the quality of the branch metric obtained based on the estimated channel impulse response is determined to be good, and a normal signal is output. Conversely, if the input signal is larger than the threshold level, it is determined that the quality of the branch metric is poor, and an alarm signal is output. The output of the comparator group 403 thus obtained is output to the output terminal group 405 and also to the comparator control circuit 404. The comparison circuit control circuit 404 controls the comparator that has output the alarm signal once so that the alarm signal is always output. This control is performed to remove the influence from the quality of the branch metric because the channel impulse response estimating circuit 1001 may diverge once the channel impulse response estimation fails. As described above, the quality of the branch metric of each diversity branch is output from the output terminal 405, respectively.

The output from the branch metric quality estimation circuit 1003 is
Output to the branch metric synthesis circuit 1004. The branch metric combining circuit 1004 combines the branch metrics in each diversity branch based on the branch metric quality in each diversity branch obtained from the branch metric quality estimating circuit 1003, and outputs the result as a branch metric for the entire diversity. The branch metric synthesis circuit can be configured, for example, as shown in FIG.

In FIG. 5, a signal from the output terminal group 306 in FIG. 3 or the signal from the output terminal group 406 in FIG. Further, a branch metric in each diversity branch provided as an output of the branch metric operation circuit group 1002 is input from an input terminal 501. The signals input from the input terminal group 500 are input to the signal conversion circuit group 502, respectively. In each circuit of the signal conversion circuit group 503,
If the input signal is a normal start of operation, it is converted to 1 and if it is an alarm signal, it is converted to 0 and output. Signal conversion circuit group 50
The output of 2 is input to an addition circuit 502 and added, and a divider 506
Is output to At this time, the output of the adding circuit 504 becomes equal to the number of signal conversion circuits to which a normal signal is input. On the other hand, the output of the signal conversion circuit group 502 is
Also supplied in 05. The multiplier group 505 multiplies the branch metric obtained from the k-th diversity branch by the output of the circuit to which the quality of the branch metric of the k-th diversity branch among the signal conversion circuit group 502 is input. The multiplication result is added by the adder 503. The output of the adder 503 is obtained by a divider 506.
The result is divided by the value obtained from the adder 504 and output to the output terminal 507 as a branch metric of the entire diversity.

The branch metric of the entire diversity obtained from the branch metric synthesis circuit is input to a soft-decision Viterbi demodulation circuit 1005 that can realize maximum likelihood sequence demodulation, and obtains a determination result. This soft-decision Viterbi demodulation circuit uses ACS (Add, Compa
ordinary Viterbi decoder consisting of a re and select) circuit and a path memory (for example, Suzuki, Tajima, "Realization of Maximum Likelihood Decoder for Convolutional Codes", IEICE Transactions A, Vo)
l. J73-A, No. 2 pp. 225-231, February 1990).

(Effects of the Invention) According to the present invention, when data transmission is performed via a communication channel in which intersymbol interference with time variation occurs, intersymbol interference can be efficiently removed and data transmission characteristics can be improved.

[Brief description of the drawings]

FIG. 1 is a system diagram showing an embodiment in which the adaptive diversity receiver according to the present invention is applied to a receiver having L diversity branches. In the figure, 1000
L is a diversity branch group, 1001 is L L channel impulse response estimation circuits, 1002 is L branch metric calculation circuits, 1003 is a branch metric quality estimation circuit, 1004 is a branch metric synthesis circuit, 1005 is a soft-decision Viterbi demodulation circuit, and 1006 is an output terminal. FIG. 2 is a system diagram showing an embodiment of the communication channel impulse response estimation circuit of FIG. In the figure, 200 and 201 are input terminals, 202 is M tap coefficients, 203 is M-1 registers, 204 is M multiplier groups, 205 is an adder, 206 is a subtractor, 207 is a processor, 208 is a delay circuit, 210 is an output terminal group, and 211 is an output terminal. FIG. 3 is a system diagram showing one embodiment of the branch metric quality calculation circuit of FIG. In the figure, 300 is an L input terminal group, 302 is an L power detection circuit group, 303 is an L comparator group, 304 is a comparator control circuit, and 305 is an L output terminal group. FIG. 4 is a system diagram showing another embodiment of the branch metric quality operation circuit of FIG. In the figure, 400 is L
× M input terminal group, 401 is L delay circuit group, 402 is L communication channel impulse response variation operation circuit group, 403 is L comparator group, 404 is comparator control circuit, 405 is 1 Output terminal. FIG. 5 is a system diagram showing an embodiment of the branch metric synthesis circuit of FIG. In the figure, 500 is an L input terminal group, 501 is an L input terminal group, 502 is an L signal conversion circuit group, 503 and 504 are adders, 505 is an L multiplier group, and 506
Is a divider, and 507 is an output terminal. FIG. 6 is an example of a diversity receiver using a conventional adaptive MLSE receiver. In the figure, 600 is a diversity antenna, 601 is an adder, 602 is a communication channel impulse response estimation circuit, 603 is a branch metric operation circuit, 604 is a maximum likelihood sequence estimation circuit, and 605 is an output terminal.

Claims (5)

(57) [Claims] 1. A receiving apparatus for performing diversity reception using a plurality of (L) antennas, comprising: (a) connected to each diversity branch, receiving each reception signal in each diversity branch and a determination result to be described later, L number of channel impulse responses for estimating the channel impulse response to each of the diversity branches and outputting the estimation result of the channel impulse response to each of the diversity branches and the internal state used to obtain the estimation result Estimating circuits,
(B) connected to each of the diversity branches, inputting the received signal in each of the diversity branches and the estimation result of the communication channel impulse response to each of the diversity branches, and obtaining a branch metric of the received signal in each of the diversity branches And (c) estimating a branch metric quality for a received signal in each of the diversity branches based on the input signal, using an output from the L channel impulse response estimating circuits as an input signal. And (d) the outputs of the L branch metric calculation circuits and the outputs of the branch metric quality calculation circuits, and a branch for the received signal in each of the diversity branches. A branch metric synthesizing circuit that synthesizes a branch metric for a received signal in each of the diversity branches based on a trick quality and outputs the result as a branch metric of the entire diversity; and (e) receives the branch metric of the entire diversity as an input and outputs the diversity. An adaptive diversity reception system, comprising: a maximum likelihood sequence estimation circuit that performs maximum likelihood sequence estimation based on the entire branch metric and outputs a determination result. 2. Each of the L communication channel impulse response estimation circuits includes: (a) receiving the determination result as an input, outputting an input signal to each tap to an adaptive control processor described later, Each tap coefficient at i
_k (i) (k = 1, 2,..., L) a transversal filter composed of M taps for inputting / outputting to / from an adaptive control processor, and (b) receiving a received signal in each of the diversity branches, A delay circuit that delays by the demodulation time generated in the maximum likelihood sequence estimation circuit,
(C) An output of the delay circuit and an output of the transversal filter are input signals, and an error signal of the input signal is output to an adaptive control processor and an output terminal as an internal state used to obtain the estimation result. And (d) an input result of the subtraction circuit, an input signal to each tap, and each tap coefficient, and an estimation result of estimating a channel impulse response to each diversity branch at time i + 1. The adaptive diversity receiving apparatus according to claim 1, further comprising: an output terminal; and an adaptive control processor that outputs _k (i + 1) as the tap coefficients to the transversal filter. 3. The branch metric quality operation circuit includes:
(A) L output from the channel impulse response estimation circuit connected to each of the diversity branches
(B) L signal power detection circuit groups for inputting the error signals and obtaining the signal power of the input signal; and (b) inputting the respective outputs of the L signal power detection circuit groups, If the level is lower than a predetermined level, a normal signal is output, if the level is higher, an alarm signal is output.If the alarm signal is output once by a control signal from a comparator control signal, the L comparators controlled to continue to output an alarm signal;
(C) inputting signals output from the L comparators,
3. The adaptive diversity receiver according to claim 2, further comprising: a comparator control circuit that outputs the control signal to the comparator that has received the alarm signal once. 4. The branch metric quality operation circuit comprises:
(A) An estimation result _k (i) (k = 1, 2,... L) of estimating a channel impulse response up to each of the diversity branches connected to each of the diversity branches
And (b) output _k (i) of each of the L delay circuit groups and the respective diversity at time i + 1 The estimation result _k (i + 1) _obtained by estimating the channel impulse response up to the branch is input,
Difference vector delta _k = _k of the input signal vector _k (i) and _{k (i + 1) (i} + 1) - seeking k (i), the absolute value of each element of the difference vector delta _k, the absolute value of each element L communication channel impulse response variation calculation circuit groups that output the largest one among them;
The output of each of the communication channel impulse response fluctuation calculation circuits is input. If the input signal is lower than a predetermined level, a normal signal is output. If the input signal is higher, an alarm signal is output. When the alarm signal is output once by the control signal from the device control signal, L comparators controlled to continue outputting the alarm signal; and (d) the L comparators 3. The adaptive diversity receiver according to claim 2, further comprising: a comparator control circuit that receives the output signal and outputs the control signal to the comparator that has output the alarm signal once. apparatus. 5. The branch metric synthesis circuit according to claim 1,
(A) L which receives each output signal from the L comparators and outputs 1 when the input signal is equal to the normal signal and outputs 0 when the input signal is equal to the alarm signal
(B) a first adder circuit that inputs and adds the respective outputs of the L signal conversion circuit groups, and (c) each of the L signal conversion circuit groups. An L number of multiplication circuit groups for multiplying an output and each of the branch metrics obtained from each diversity branch according to claim 1 connected to the signal conversion circuit group,
(D) a second addition circuit for adding the outputs of the L multiplication circuit groups, and (e) an input of the output of the addition circuit 1 and an output of the second addition circuit, 5. The adaptive diversity receiving apparatus according to claim 3, further comprising a dividing circuit for dividing an output of the adding circuit by an output of the first adding circuit.