JP4611271B2 - Receiver - Google Patents

Description

  The present invention relates to a receiving apparatus that demodulates signals received by a plurality of antennas, and more particularly to a receiving apparatus that performs demodulation using a maximum likelihood determination method.

  There is a maximum likelihood determination method as one of data determination methods in digital communication. A receiver using the maximum likelihood determination method (hereinafter referred to as a receiver) creates a replica based on a transmission path response and a transmission symbol candidate (candidate combination of bit values of transmission symbols), and the difference between the received signal and the replica Compute a metric that represents At this time, replicas are created for all possible transmission symbol candidates, and metrics are calculated. Then, a transmission symbol candidate corresponding to the replica having the smallest metric is output as a determination result. This maximum likelihood determination method has excellent reception performance, but generally requires a large amount of calculation because it calculates metrics for all possible combinations of transmission bits.

  As a technique for reducing the calculation amount, for example, as described in Non-Patent Document 1 below, a calculation amount reduction type maximum likelihood determination method has been proposed. According to the following Non-Patent Document 1, the amount of calculation required for maximum likelihood determination is reduced by limiting the transmission symbol candidates for calculating the metric to only those within a certain distance from the received signal. .

  On the other hand, the computational complexity reduction type maximum likelihood determination method has a problem that it is difficult to calculate a soft decision value used for error correction decoding. In general, the soft decision value in units of bits is obtained by a log likelihood ratio between the likelihood when the bit is “−1” and the likelihood when the bit is “+1”. However, the computation amount reduction type maximum likelihood determination method does not calculate metrics for all transmission symbol candidates. Therefore, there is a possibility that there is no symbol including “−1” or “+1” in the transmission symbol candidates for calculating the metric. In such a case, the soft decision value cannot be calculated.

  Further, as a technique for solving the problem of the above-described calculation amount reduction type maximum likelihood determination method, for example, there is a technique described in Patent Document 1 below. In the technique described in Patent Document 1 below, when a metric corresponding to either “−1” or “+1” is not calculated in the maximum likelihood determination method of a calculation amount reduction type, it is determined in advance from average noise power or the like. The calculated threshold is used as a substitute for the uncalculated metric to calculate a soft decision value.

JP 2006-50432 A Emanuele Viterbo, Joseph Boutros, “A Universal Lattice Code Decoder for Fading Channels”, IEEE Transactions on Information Theory, Vol.45, No.5, pp.1639-1642, July 1999.

  However, according to the above-described conventional technique, when calculating the soft decision value in the maximum likelihood determination method of the calculation amount reduction type, the threshold value determined in advance from the average noise power or the like is used instead of the metric that is not calculated. There is a problem that the accuracy of the value is not sufficient, and as a result, good decoding characteristics cannot be obtained.

  The present invention has been made in view of the above, and it is possible to obtain a soft decision value with high accuracy and to provide good communication in a maximum likelihood determination method of a calculation amount reduction type. The object is to obtain a device.

  In order to solve the above-described problems and achieve the object, the present invention provides a receiving apparatus that demodulates signals received by a plurality of antennas (referred to as received signals) using a maximum likelihood determination method, First signal determining means for estimating a hard decision sequence based on a signal and a transmission path response of the received signal, and outputting the hard decision sequence and first likelihood information corresponding to the hard decision sequence; A new signal point obtained by excluding a signal point at which the bit value of the target bit position matches the bit value of the same bit position in the hard decision sequence from all possible signal points of the received signal for each bit position of the symbol Using the signal point information generating means for generating an arrangement and the new signal point arrangement generated for each bit position of the transmission symbol, signal determination based on the received signal and the transmission path response is performed, and the signal determination is performed. A second signal determining means for calculating and outputting second likelihood information corresponding to the result; and a soft decision for calculating a soft decision value using the first likelihood information and the second likelihood information. And a calculating means.

  According to the present invention, for each bit position of the transmission symbol, signal determination is performed using a signal point arrangement excluding signal points where the bit value of the bit position matches the bit value of the bit position of the hard decision sequence. Therefore, the likelihood information of the bit opposite to that of the hard decision sequence is calculated, so that even when performing the maximum likelihood determination of the calculation amount reduction type, a highly accurate soft decision value can be obtained and good communication can be performed. There is an effect that it can be provided.

  Embodiments of a receiving apparatus according to the present invention will be described below in detail with reference to the drawings. Note that the present invention is not limited to the embodiments.

Embodiment 1 FIG.
FIG. 1 is a diagram illustrating a configuration example of a first embodiment of a receiving device (hereinafter referred to as a receiver) according to the present invention. The receiver according to the present embodiment includes antennas 1-1 to 1-N (N is the number of antennas), analog signal processing units 2-1 to 2-N, and A / D (analog / digital) conversion unit 3. -1 to 3-N, a digital signal demodulator 4 that performs a characteristic operation of the present invention, a deinterleaver 5, and a decoder 6.

  The overall operation of the receiver of this embodiment will be described with reference to FIG. First, the antennas 1-1 to 1-N receive high-frequency analog signals (modulated signals). The analog signal processing units 2-1 to 2-N perform analog signal processing such as down-conversion on the signals received by the antennas 1-1 to 1-N, and A / D conversion units 3-1 to 3-N. Output to. Next, the A / D conversion units 3-1 to 3 -N convert the analog signals input from the analog signal processing units 2-1 to 2 -N into baseband digital signals and output them to the digital signal demodulation unit 4. The digital signal demodulator 4 performs demodulation processing described later on the input signal and outputs demodulated data to the deinterleaver 5. The deinterleaving unit 5 performs a process opposite to the interleaving performed by the transmitter on the demodulated data, and outputs it to the decoding unit 6. Then, the decoding unit 6 performs error correction decoding processing, which is soft decision decoding, on the output signal from the deinterleave unit 5, and outputs the result as a decoded bit sequence.

  In the present embodiment, it is assumed that information on the transmitter such as the interleaving method performed by the transmitter, the number of antennas of the transmitter, and the number of transmission bits is known.

  Next, an operation (demodulation process) in which the digital signal demodulator 4 demodulates the input signal, which is a feature of the present embodiment, will be described. FIG. 2 is a diagram illustrating a configuration example of the digital signal demodulation unit 4. The digital signal demodulator 4 includes a first signal determination unit 41, a transmission path estimation unit 42, a signal point information generation unit 43, a second signal determination unit 44, and a soft decision calculation unit 45. Yes.

  First, the N baseband digital signals which are outputs from the A / D conversion units 3-1 to 3-N are input to the digital signal demodulation unit 4. Hereinafter, for convenience of explanation, these input signals are expressed by N-dimensional column vectors and are referred to as “reception signal vectors”. When the digital signal demodulator 4 receives the received signal vector from the A / D converters 3-1 to 3 -N, first, the transmission path estimator 42 estimates the transmission path response of the desired signal based on the received signal vector. To do. The channel response estimation method is not particularly limited. For example, a known sequence (such as a pilot signal) created to be orthogonal between transmission antennas is transmitted in advance, and the known sequence data and the known sequence are transmitted. It is conceivable to calculate the transmission line response by the least square method using the received signal of the data.

  When the transmitter transmits M different signals simultaneously from the M (M is an integer greater than or equal to 1) element antenna using the same frequency, the transmission path response uses M and the number of receiving antennas N of the receiver. It can be represented by a matrix of N rows and M columns. Note that the transmission line response may be a vector or a scalar. In this case, the value of M or N (the number of elements in the row or column) may be considered as 1, and in this description, the transmission is also included. The path response is called “transmission path response matrix”. Therefore, the estimation result of the transmission path response of the transmission path estimation unit 42 is also referred to as a transmission path estimation matrix (transmission path response matrix estimation result). Further, the present invention can be applied as it is even when the transmission line response matrix is a vector or a scalar.

  Next, the first signal determination unit 41 transmits based on the reception signal vector output from the A / D conversion units 3-1 to 3-N and the transmission path estimation matrix output from the transmission path estimation unit 42. Symbol estimation is performed to obtain a hard decision sequence as an estimation result. Further, the first signal determination unit 41 calculates likelihood information α corresponding to the hard decision series. Then, the first signal determination unit 41 sends the estimated hard decision sequence to the signal point information generation unit 43 and the soft decision calculation unit 45, and likelihood information α corresponding to the hard decision sequence to the soft decision calculation unit 45, respectively. Output.

  As the estimation process of the first signal determination unit 41, for example, any signal determination technique suitable for the assumed communication system can be used, such as a calculation amount reduction type maximum likelihood determination method such as “sphere decoding”. It is. In this embodiment, a metric based on a square error between a received signal vector and a replica vector generated from a transmission path estimation matrix and a transmission symbol candidate is used as likelihood information. However, the present invention is not limited to this. Any likelihood information can be used as long as it is input to the soft decision.

  Next, based on the hard decision sequence estimated by the first signal determination unit 41, the signal point information generation unit 43 does not include the same bit value of the hard decision sequence for each bit position of the transmission symbol. A new signal point arrangement is generated and output to the second signal determination unit 44.

  The processing of the signal point information generation unit 43 will be described in detail with reference to FIGS. FIG. 3 shows an example of a new signal point arrangement generated by the signal point information generation unit 43 when the hard decision series that is the output of the first signal determination unit 41 is “1-1, 11”. ing. Specifically, in FIG. 3, 31 in FIG. 3-1 is a signal point arrangement that does not include “1” that is the first bit of the hard decision sequence in the first bit, and 32 in FIG. 3-2 is the second bit. Signal point arrangement that does not include "-1" that is the second bit of the hard decision sequence in the eye, and 33 in FIG. 3-3 includes "1" that is the third bit of the hard decision sequence in the third bit No signal point arrangement, 34 in FIG. 3-4 shows a signal point arrangement that does not include “1” that is the fourth bit of the hard decision sequence in the fourth bit.

  In this example, the signal received by the receiver according to the present embodiment is a signal transmitted from a transmitter with M = 2 that performs QPSK modulation (corresponding to a total of 4-bit transmission with 2 bits for each antenna). . When the symbol transmitted by the first transmitting antenna is symbol # 1, and the symbol transmitted by the second transmitting antenna is symbol # 2, the hard decision sequence is expressed as ““ first bit of symbol # 1 ”“ symbol # 1 ”. "The second bit of 1", "the first bit of symbol # 2", "the second bit of symbol # 2", and hereinafter, the first bit of symbol # 1 is the first bit The second bit of symbol # 1 is called the second bit, the first bit of symbol # 2 is called the third bit, and the second bit of symbol # 2 is called the fourth bit. The signal point arrangement used by the signal determination unit 41 is four points of “11”, “1-1”, “−11”, and “−1−1” for both the symbol # 1 and the symbol # 2.

  For example, in the present embodiment, since the first bit of the hard decision sequence is “1”, the signal point information generation unit 43 has a signal point whose first bit includes “1”, that is, symbol # 1. A new signal point arrangement (31 in FIG. 3-1) is generated by deleting all the signal points of “11” and “1-1” from 4 signal points, and a new signal point for the first bit is generated. Output as arrangement. Similarly, 32 of FIG. 3-2, 33 of FIG. 3-3, and 34 of FIG. 3-4 are output as new signal point arrangements from the second bit to the fourth bit, respectively.

Next, the second signal determination unit 44 receives as input the new signal point arrangement for the number of transmission bits generated by the received signal vector, the transmission path estimation matrix, and the signal point information generation unit 43, and performs signal determination and likelihood information. β _i (i is a corresponding transmission bit. When j is the number of transmission bits, i is an integer from 1 to j). Whereas the first signal determination unit 41 performs estimation using a signal arrangement including all signal points, the second signal determination unit 44 uses a signal point information generation unit for each transmission bit. Assuming that the signal is transmitted using the new signal point arrangement generated at 43, signal determination is performed for each signal point of the signal point arrangement to estimate a transmission symbol. Then, likelihood information β _i corresponding to the estimation result (for each transmission bit) is calculated, and the likelihood information β _i is output to the soft decision calculation unit 45.

  As described above, the new signal point arrangement generated by the signal point information generation unit 43 is generated so that each bit of the transmission bit does not include the bit value that is the same as the bit of the hard decision sequence. ing. Therefore, the determination result of the second signal determination unit 44 is obtained as an estimation result for each bit, a sequence having bits opposite to the hard decision sequence estimated by the first signal determination unit 41.

  Note that any method can be applied to the signal determination method used in the second signal determination unit 44. For example, a method of performing signal separation using a new signal point arrangement created by the signal point information generation unit 43 after performing signal separation using a filter based on a MMSE (Minimum Mean Square Error) standard may be considered. .

Next, the soft decision calculation unit 45 is based on the likelihood information α input from the first signal determination unit 41 and the likelihood information β _i of several transmission bits input from the second signal determination unit 44. The soft decision value for each bit is calculated, and the obtained soft decision value is output to the deinterleave unit 5.

  Here, the soft decision value can be generally expressed as a log likelihood ratio between the likelihood information whose bit for which the soft decision value is calculated is “−1” and the likelihood information whose value is “+1”. . When a metric based on the square error between the received signal vector and the replica vector is used as the likelihood information, the soft decision value of the kth bit is a metric corresponding to “−1” as shown in the following equation (1). And the metric corresponding to “+1” can be easily calculated.

(Soft decision value of the kth bit)
= (Metric corresponding to “−1”) − (Metric corresponding to “+1”) (1)

Of the right side of Equation (1), the first term and the second term correspond to either the likelihood information α or the likelihood information β _i of the present embodiment. For example, in the example used in the description of FIG. 3 described above, the soft decision value for each bit can be calculated using the following equations (2) to (5).

(Soft decision value of the first bit)
= (Likelihood information β ₁ estimated using 31 in FIG. 3-1) − (likelihood information α) (2)

(Soft decision value of the second bit)
= (Likelihood information α) − (likelihood information β ₂ estimated using 32 in FIG. 3-2) (3)

(Soft decision value of the third bit)
= (Likelihood information β ₃ estimated using 33 in FIG. 3-3) − (likelihood information α) (4)

(Soft decision value of the 4th bit)
= (Likelihood information β ₄ estimated using 34 in FIG. 3-4) − (likelihood information α) (5)

  Next, the deinterleaving unit 5 performs a process reverse to the interleaving performed by the transmitter on the soft decision value calculated by the above procedure, and outputs the result to the decoding unit 6. The decoding unit 6 performs soft input decoding using the input signal. In this way, the final decoded signal is obtained.

In the present embodiment, the signal point information generation unit 43 generates a number of new signal point arrangements equal to the number of transmission bits, and the second signal determination unit 44 that receives the new signal point arrangements estimates the number of transmission bits. The degree information β _i is calculated, but among the transmission bits, the first signal determination unit 41 has already calculated both the “+1” metric and the “−1” metric. Alternatively, a soft decision calculation may be performed using these already calculated metrics. Thereby, the amount of calculation can be reduced.

  In the present embodiment, the digital signal demodulation unit 4 includes the first signal determination unit 41 and the second signal determination unit 44 as signal determination units. May be integrated into one. For example, in the aforementioned “sphere decoding”, the process of performing signal separation using a filter of ZF (Zero Forcing) nor MMSE (Minimum Mean Square Error) norm is included in the process of performing the maximum likelihood determination of the calculation amount reduction type. Yes. For this reason, the signal determination process corresponding to the second signal determination unit 44 may be replaced with these processes included in “sphere decoding” of the first signal determination unit 41.

As described above, in the present embodiment, for each bit position of the transmission symbol, a signal point arrangement excluding a signal point in which the bit value of the bit position matches the bit value of the bit position of the hard decision sequence is used. By performing signal determination, likelihood information (the likelihood information β _i ) of bits opposite to the hard decision series is calculated. As a result, even when the calculation amount reduction type maximum likelihood determination is performed, it is possible to generate a highly accurate soft decision value and to provide good communication.

Embodiment 2. FIG.
FIG. 4 is a diagram illustrating a configuration example of the digital signal demodulation unit 4a according to the second embodiment of the receiver according to the present invention. The configuration other than the digital signal demodulator 4a of the present embodiment is the same as that of the first embodiment. Hereinafter, a different part from Embodiment 1 is demonstrated.

  As shown in FIG. 4, the digital signal demodulator 4 a of the present embodiment includes a second signal determination unit 44 a instead of the second signal determination unit 44 of the digital demodulation unit 4 of the first embodiment. Components having the same functions as those of the first embodiment are denoted by the same reference numerals and description thereof is omitted.

  FIG. 5 is a diagram illustrating a configuration example of the second signal determination unit 44a in the digital signal demodulation unit 4a of the present embodiment. As shown in FIG. 5, the second signal determination unit 44 a includes a symbol determination unit 441, a replica subtraction unit 442, and a third signal determination unit 443.

  The second signal determination unit 44a in the present embodiment, when the received signal is composed of a plurality of symbols, includes a symbol including a bit for calculating a soft decision value (referred to as the i-th bit) and other symbols. The symbols including the bits for calculating the soft decision value are determined first, and then the remaining symbols are determined.

  Hereinafter, the signal determination procedure of the second signal determination unit 44a of the present embodiment will be described. First, the symbol determination unit 441 calculates a soft decision value from the new signal point arrangement generated by the signal point information generation unit 43 based on the bit position for calculating the soft decision value and the value of the bit. Only signal points of symbols including symbols to be transmitted (symbols transmitted from the same transmission antenna) are selected.

  Next, based on the received signal vector and the transmission path estimation matrix, determination is performed using the selected signal point using arbitrary signal determination means. That is, the determination is made by reducing the number of signal points of the signal point arrangement generated in the signal point information generation unit 43. For example, in the example shown by 31 in FIG. 3-1, when calculating the soft decision value of the first bit, the symbol # 1 is selected (two signal points are selected), and the signal decision is performed.

  First, the symbol determination unit 441 uses the transmission path estimation matrix and the received signal vector to separate the symbol component selected from the received signal vector into other components based on, for example, a method using a filter based on the MMSE standard. . Then, signal determination is performed by the same method as the processing of the second signal determination unit 44 of the first embodiment using the separated selected symbol component and the transmission path estimation matrix, and the determination value (soft determination value is determined). (Estimated value of transmission symbol including bits to be calculated) is output to replica subtraction unit 442.

  Next, replica subtraction section 442 creates a replica of the symbol using the determination value output from symbol determination section 441 and the transmission path estimation matrix, and subtracts the replica from the received signal vector. Then, the subtraction result is output to the third signal determination unit 443.

Next, the third signal determination unit 443 performs signal determination on symbols other than the symbols determined by the symbol determination unit 441 based on the signal output from the replica subtraction unit 442 and the transmission path estimation matrix, and determines the determination result. The corresponding likelihood information β _i is calculated. Here, the signal point arrangement used for the determination may be a new signal point arrangement generated by the signal point information generation unit 43 or a normal signal point arrangement used for transmission. Also, any method used for signal determination can be applied.

Likelihood information β _i corresponding to all bits is obtained by repeating the above processing for the number of transmission bits. Processing other than the calculation of the likelihood information β _i is the same as in the first embodiment.

  As described above, in the present embodiment, the second signal determination unit 44a first determines a symbol including a bit that is a target of soft decision value calculation, and receives a replica calculated from the determination result as a received signal. After removing from the vector, the remaining signal determination is performed, and the symbol including the bits for which the soft decision value is calculated is determined in a state where the number is smaller than the normal signal point arrangement. For this reason, it is considered that determination of a symbol including a bit that is a target of soft decision value calculation has fewer errors and higher accuracy than a case where determination is performed including other signal points. Further, since the remaining symbols are determined after removing the influence of the signal estimated with high accuracy, it is possible to obtain highly accurate likelihood information as a whole.

Embodiment 3 FIG.
FIG. 6 is a diagram illustrating a configuration example of the digital signal demodulating unit 4b according to the third embodiment of the receiver according to the present invention. The configuration other than the digital signal demodulator 4b of the present embodiment is the same as that of the first embodiment. Hereinafter, a different part from Embodiment 1 is demonstrated.

  As shown in FIG. 6, the digital signal demodulator 4b according to the present embodiment includes a second signal determination unit 44b instead of the second signal determination unit 44 of the digital demodulation unit 4 according to the first embodiment. Components having the same functions as those of the first embodiment are denoted by the same reference numerals and description thereof is omitted.

  FIG. 7 is a diagram illustrating a configuration example of the second signal determination unit 44b in the digital signal demodulation unit 4b according to the present embodiment. As shown in FIG. 7, the second signal determination unit 44b includes a bit determination unit 444, a bit probability calculation unit 445, an expected value calculation unit 446, and a likelihood information calculation unit 447. In the present embodiment, likelihood information calculation using the expected value of the symbol is executed in the second signal determination unit 44b.

  Hereinafter, the process of the second signal determination unit 44b of the present embodiment will be described. First, the bit determination unit 444 of the second signal determination unit 44b is based on the received signal vector, the transmission path estimation matrix, the new signal point arrangement generated by the signal point information generation unit 43, and the bit for which the soft decision value is to be calculated. Estimates a bit sequence different from the hard decision sequence, and calculates a soft decision value for the bit sequence. This process is performed on a new signal point arrangement of several transmission bits, and sets of soft decision values (soft decision sequences) corresponding to the number of transmission bits are obtained. At this time, since it is not necessary to calculate the probability for the bit position obtained by inverting the bit value of the hard decision value (because the bit value is fixed), the soft decision value is not calculated. As the processing of the bit determination unit 444, an arbitrary soft determination method can be applied. For example, it can be realized by performing signal separation using a filter based on the MMSE standard. In addition, when the calculation amount reduction type maximum likelihood determination method is used as the processing of the bit determination unit 444, the same problem as the problem of the present invention may occur. Even in such a case, for example, by repeatedly applying the processing of the present invention inside the bit determination unit 444, it is possible to calculate a soft decision value for a bit sequence in which a bit for which a soft decision value is to be calculated is different from the hard decision sequence. is there. Then, the estimated soft decision value is output to bit probability calculation section 445.

  The bit probability calculation unit 445 calculates the probability that each bit is “−1” and the probability that it is “+1” using the input soft decision value. For example, when the soft decision value calculation using the log likelihood ratio is performed by the bit decision unit 444, the soft decision value for the kth bit is represented by L, and the following equation (6), It can be calculated in (7).

(Probability that the kth bit is “−1”) = e ^L / (e ^L +1) (6)

(Probability that the k-th bit is “+1”) = 1 / (e ^L +1) (7)

  Next, the bit probability calculation unit 445 outputs the calculation result of the probability that each bit is “−1” and the probability that the bit is “+1” to the expected value calculation unit 446.

  The expected value calculation unit 446 calculates the expected value of the transmission symbol based on the input probability and the new signal point arrangement.

  Here, how to obtain the expected value will be described taking the signal point arrangement 31 shown in FIG. 3A as an example. In symbol # 1 of 31 in FIG. 3A, there are two signal points corresponding to “−11” and signal points corresponding to “−1-1”. When calculating the expected value for the symbol # 1, first, the probability that the signal point “-11” appears and the probability that the signal point “−1-1” appears are calculated. However, in this case, since the first bit is determined by a signal point arrangement fixed to “−1” (corresponding to probability 1), what is necessary for the calculation is the probability that “1” appears in the second bit. And the probability that “−1” appears in the second bit. After that, each probability is multiplied by the corresponding signal point coordinate value (corresponding to a coordinate value when the value of each bit is a two-dimensional coordinate having two axes), and the expected value is calculated. Similarly, the expected value for symbol # 2 is also multiplied by the signal point coordinate value corresponding to the probability that each of the signal points "-1-1", "-11", "1-1", "11" appears, Calculate by adding together.

  The expected value of the symbol calculated here does not necessarily coincide with the signal point on the new signal point arrangement, and may indicate a position shifted from the signal point. For example, the expected value for symbol # 1 of 31 in FIG. 3-1 is “−1−” when the probabilities of “−11” and “−1−1” are 0.3 and 0.7, respectively. 0.4 ".

  The above processing is performed for the number of transmission bits, and the expected value of the transmission symbol is output to likelihood information calculation section 447.

The likelihood information calculation unit 447 calculates a metric between the received signal vector and the expected value of the replica obtained by multiplying the expected value of the symbol calculated by the expected value calculation unit 446 by the transmission path estimation matrix, and the likelihood. Information β _i . Processing other than the calculation of the likelihood information β _i is the same as in the first embodiment.

As described above, in the present embodiment, the second signal determination unit 44b uses the likelihood information (corresponding to the likelihood information β _i ) of the opposite bit value of each bit of the hard decision sequence as the expected value of the symbol. It was decided to calculate using This makes it possible to minimize errors when calculating likelihood information from erroneous symbol determination results when estimating likelihood information of the opposite bit value of each bit of the hard decision sequence. This contributes to high-precision soft decision calculation.

Embodiment 4 FIG.
FIG. 8 is a diagram illustrating a configuration example of the digital signal demodulating unit 4c according to the fourth embodiment of the receiver according to the present invention. The configuration other than the digital signal demodulator 4c of the present embodiment is the same as that of the first embodiment. Hereinafter, a different part from Embodiment 1 is demonstrated.

  As shown in FIG. 8, the digital signal demodulator 4c of the present embodiment includes a first signal unit 41a instead of the first signal determination unit 41 of the digital signal demodulator 4 of the first embodiment. The likelihood information correction unit 46 is provided. Components having the same functions as those of the first embodiment are denoted by the same reference numerals and description thereof is omitted.

  FIG. 9 shows a configuration example of the first signal determination unit 41a. The first signal determination unit 41 a includes two signal determination units, a fourth signal determination unit 411 and a fifth signal determination unit 412, and a weight coefficient calculation unit 413. In the present embodiment, an error due to a difference in determination method between the first signal determination unit 41a and the second signal determination unit 44 is corrected.

  The first signal determination unit 41a and the second signal determination unit 44 may perform the determination using the same method. However, in general, the first signal determination unit 41a performs a highly accurate determination, and the second signal determination unit 41a performs the second determination. It is expected that the signal determination unit 44 makes a determination with priority given to speed. In that case, the accuracy improvement technique by correction as in the present embodiment is effective.

  Hereinafter, the process of the first signal determination unit 41a will be described. First, the fourth signal determination unit 411 of the first signal determination unit 41a estimates the hard decision sequence based on the received signal vector that is the input signal and the transmission path estimation matrix received from the transmission path estimation unit 42. The likelihood information α corresponding to the hard decision series is calculated.

  Further, the fifth signal determination unit 412 uses the same method as the signal determination method used in the second signal determination unit 44, and receives the received signal vector as the input signal and the transmission path estimation received from the transmission path estimation unit 42. Based on the matrix, the hard decision series and the likelihood information θ are calculated. For example, a computational complexity reduction type maximum likelihood determination method such as “sphere decoding” is applied to the fourth signal determination unit 411, and signal determination by a filter based on the MMSE standard is applied to the fifth signal determination unit 412. Apply.

  Then, the fourth signal determination unit 411 outputs the obtained hard decision series to the soft decision calculation unit 45 and the likelihood information α to the soft decision calculation unit 45 and the weight coefficient calculation unit 413. The fifth signal determination unit 412 outputs the obtained likelihood information θ to the weight coefficient calculation unit 413.

  Next, the weighting factor calculation unit 413 calculates a weighting factor from the input likelihood information α and θ, and outputs the weighting factor to the likelihood information correction unit 46. The two pieces of likelihood information α and θ are likelihood information calculated by the fourth signal determination unit 411 and the fifth signal determination unit 412 using different signal determination methods for the same signal, and α and The difference in the value of θ represents the difference in signal determination accuracy between the two signal determination methods. The weighting coefficient calculation unit 413 calculates a weighting coefficient for correcting the accuracy difference between the two pieces of likelihood information. For example, as the weighting coefficient, a ratio of likelihood information α and θ is calculated, and a value averaged between a plurality of symbols is used.

The likelihood information correction unit 46 corrects the likelihood information β _i using the likelihood information β _i output from the second signal determination unit 44 and the weighting factor output from the first signal determination unit 41a. The correction value β _i ′ is output to the soft decision calculation unit 45. The likelihood information β _i is corrected by an appropriate method so as to correct the error in accordance with the weighting factor calculation method. For example, when the average value of the ratio of the likelihood information α and θ described above is used as a weighting coefficient, a method of multiplying the weighting coefficient and the likelihood information β _i to obtain corrected likelihood information can be considered.

  The processes other than the first signal determination unit 41a and the likelihood information correction unit 46 are the same as those in the first embodiment.

  Thus, in the present embodiment, the accuracy difference existing between the likelihood information corresponding to the hard decision sequence and the likelihood information of the opposite bit value for each bit of the hard decision sequence, The correction is made using a weighting coefficient. Thereby, the accuracy of the soft decision value is improved, and as a result, the communication quality can be improved.

  In the present embodiment, the configuration is such that all the signal determination units exist independently. However, in the signal determination method such as “sphere decoding” described above, for example, the processing of maximum likelihood determination with a reduced amount of computation is performed. The process includes signal separation by a filter based on the ZF norm and MMSE norms. For this reason, when the signal determination method such as “sphere decoding” is adopted in the fourth signal determination unit 411, the fifth signal determination unit 412 or the like using the signal separation function based on the ZF norm and MMSE norm included therein. The function of the second signal determination unit 44 can also be realized. For example, any two or all of the fourth signal determination unit 411, the fifth signal determination unit 412, and the second signal determination unit 44 described in FIGS. 8 and 9 are set as one signal determination unit. It is also possible to take a configuration.

Embodiment 5 FIG.
FIG. 10 is a diagram illustrating a configuration example of a fifth embodiment of a receiver according to the present invention. In the present embodiment, a digital signal demodulator 4d and a decoder 6a are provided instead of the digital signal demodulator 4 and the decoder 6 of the first embodiment. Furthermore, in this embodiment, an interleave unit 8 and adders 7 and 9 are added. Components having the same functions as those of the first embodiment are denoted by the same reference numerals and description thereof is omitted. Hereinafter, a different part from Embodiment 1 is demonstrated.

  In the present embodiment, deinterleaving unit 5, decoding unit 6a, interleaving unit 8, and adders 7 and 9 operate as signal decoding means as a whole. The decoding unit 6a is a so-called soft input / soft output decoder, and the receiver according to the present embodiment generates a priori information based on the decoded data output from the decoding unit 6a, and performs digital signal demodulation based on the prior information. The unit 4d performs a signal demodulation operation. Then, after exchanging the prior information and the soft decision value a predetermined number of times between the signal decoding means and the digital signal demodulating unit 4d, the decoding unit 6a outputs the calculated decoded data as final data. In this way, it is possible to improve the determination accuracy.

  FIG. 11 is a diagram illustrating a configuration example of the digital signal demodulating unit 4d according to the present embodiment. The digital signal demodulator 4d replaces the first signal determiner 41 and the second signal determiner 44 of the digital signal demodulator 4 of the first embodiment described above with the first signal determiner 41b and the second signal determiner 41b, respectively. The signal determination unit 44c is provided. In addition, the thing of the function similar to Embodiment 1 attaches | subjects the same code | symbol, and abbreviate | omits description.

  The first signal determination unit 41b performs signal determination based on the received signal vector, the transmission path estimation matrix, and prior information that is an output signal from the interleave unit 8 described later, and a hard decision sequence and corresponding likelihood information α. Is output. For example, the first signal determination unit 41b uses “sphere decoding” or the like that performs MAP (Maximum A posteriori Probability) estimation, which is known as a method of generating likelihood information in consideration of prior information. Generate. The first signal determination unit 41 b outputs the estimated likelihood information α to the soft decision calculation unit 45 and the hard decision sequence to the signal point information generation unit 43 and the soft decision calculation unit 45.

On the other hand, the second signal determination unit 44c uses the received signal vector, the transmission path estimation matrix, the new signal point arrangement generated by the signal point information generation unit 43, and the prior information that is the output signal from the interleaving unit 8. Thus, signal determination in consideration of prior information is executed, and as a result, likelihood information β _i is obtained. As the signal determination method used here, any method can be applied as long as it can perform signal determination in consideration of prior information. For example, a method using a filter based on the MMSE norm is applicable. The likelihood information β _i obtained by the second signal determination unit 44 c is output to the soft decision value calculation unit 45.

Next, the soft decision value calculation unit 45 performs soft decision based on the likelihood information α and the hard decision series received from the first signal decision unit 41b and the likelihood information β _i received from the second signal decision unit 44c. Calculate the value.

  Next, the adder 7 that has received the soft decision value from the digital signal demodulator 4d subtracts the prior information that is the output signal of the interleave unit 8 from the soft decision value, and the result is sent to the deinterleave unit 5. Output. The deinterleaving unit 5 performs a process opposite to the interleaving performed by the transmitter on the transmission signal, and outputs the result to the decoding unit 6 a and the adder 9.

  The decoding unit 6a is realized by a so-called soft input / soft output decoder such as a SOVA (Soft Output Viterbi Algorithm) decoder or a MAP decoder, for example, for the signal received from the deinterleave unit 5 Perform decryption. The adder 9 subtracts the signal received from the deinterleaving unit 5 from the output signal (soft output) of the decoding unit 6 a and outputs the result to the interleaving unit 8. The interleaving unit 8 performs the same processing as the interleaving performed by the transmitter, and outputs the result to the first signal determination unit 41b, the second signal determination unit 44c, and the adder 7 as prior information.

  After the above process is repeated a predetermined number of times, the decoding unit 6a outputs final decoded data.

  In addition, in this Embodiment, although it was set as the structure which added the interleave part 8 etc. to the receiver concerning Embodiment 1 mentioned above, it is not restricted to this, Any one of Embodiment 2-4 mentioned above. It is good also as a structure which added the interleave part 8 etc. with respect to the receiver.

  As described above, in the present embodiment, the soft decision value and the preliminary value are transmitted between the digital signal demodulating unit 4d and the decoding unit 6a via the adder 7, the deinterleave unit 5, the adder 9, and the interleave unit 8. While exchanging information, the above-described soft decision value generation process and decoding process are repeated a predetermined number of times, and the finally obtained decoded data is output as a final result. Thereby, the accuracy of the determination data is improved, and better communication can be realized as compared with the first to fourth embodiments.

Embodiment 6 FIG.
FIG. 12 is a diagram illustrating a configuration example of a sixth embodiment of a receiver according to the present invention. In this embodiment, a digital signal demodulator 4e is provided instead of the digital signal demodulator 4 of the first embodiment, and an FFT (Fast Fourier Transform) unit 10-1 to N and a parallel-serial converter 11 are added. Has been. Components having the same functions as those of the first embodiment are denoted by the same reference numerals and description thereof is omitted. Hereinafter, a different part from Embodiment 1 is demonstrated.

  The processing of this embodiment will be described with reference to FIG. First, multicarrier high-frequency analog signals (for example, Orthogonal Frequency Division Multiplexing (OFDM), Orthogonal Frequency Division Multiple Access (OFDMA), Multi Carrier-Code Division Multiplexing (MC-CDMA) received via the antennas 1-1 to 1-N. Access) signal) is input to the FFT units 10-1 to 10-N after predetermined reception processing is performed in the analog signal processing units 2-1 to 2-N. The FFT units 10-1 to 10-N convert the time domain signal into a frequency domain signal for each subcarrier by performing discrete Fourier transform or fast discrete Fourier transform on the input signal, and digital signal demodulation is performed on the result. To the unit 4e.

  The digital signal demodulator 4e has the same configuration and function as the digital signal demodulator included in any one of the receivers of the first to fifth embodiments described above, and outputs a soft decision value according to the same procedure. However, the digital signal demodulator 4e of the present embodiment uses a frequency domain signal for each subcarrier instead of the input received signal as an input signal, and calculates a soft decision value for each subcarrier. Then, the parallel / serial converter 11 rearranges the demodulated data of each subcarrier estimated in the digital signal demodulator 4e in series, and outputs the result to the deinterleaver. Thereafter, deinterleaving section 5 and decoding section 6 perform the same processing as in Embodiment 1 for each subcarrier.

  In addition, in this Embodiment, although it was set as the structure which added FFT part 10-1 to 10-N etc. to the receiver concerning Embodiment 1 mentioned above, not only this but Embodiment 2 mentioned above. The FFT units 10-1 to 10-N may be added to any one of the five receivers, and a soft decision value may be output for each subcarrier.

  Thus, in this embodiment, when receiving a multicarrier signal in OFDM, OFDMA, or MC-CDMA, the received signal is demodulated for each subcarrier, and the above-described processing by the digital signal demodulator is applied. It was decided to. Thereby, even when OFDM, OFDMA, or MC-CDMA is employed, the same effects as those of the first to fifth embodiments can be obtained.

  As described above, the receiving apparatus according to the present invention is useful as a communication apparatus that demodulates signals received by a plurality of antennas, and is particularly suitable for communication that performs demodulation using the maximum likelihood determination method.

3 is a diagram illustrating a configuration example of a receiver in Embodiment 1. [FIG. 3 is a diagram illustrating a configuration example of a digital signal demodulating unit according to Embodiment 1. FIG. It is a figure which shows the example of the signal point arrangement | positioning which a signal point information generation part produces | generates. It is a figure which shows the example of the signal point arrangement | positioning which a signal point information generation part produces | generates. It is a figure which shows the example of the signal point arrangement | positioning which a signal point information generation part produces | generates. It is a figure which shows the example of the signal point arrangement | positioning which a signal point information generation part produces | generates. FIG. 10 is a diagram illustrating a configuration example of a digital signal demodulation unit according to a second embodiment. 10 is a diagram illustrating a configuration example of a second signal determination unit of the second embodiment. FIG. FIG. 10 is a diagram illustrating a configuration example of a digital signal demodulation unit according to a third embodiment. FIG. 10 is a diagram illustrating a configuration example of a second signal determination unit according to the third embodiment. FIG. 10 is a diagram illustrating a configuration example of a digital signal demodulation unit according to a fourth embodiment. FIG. 10 is a diagram illustrating a configuration example of a first signal determination unit according to the fourth embodiment. FIG. 11 is a diagram illustrating a configuration example of a receiver in a fifth embodiment. FIG. 10 is a diagram illustrating a configuration example of a digital signal demodulation unit according to a fifth embodiment. 69 is a diagram illustrating a configuration example of a receiver in Embodiment 6. [FIG.

Explanation of symbols

1-1 to 1-N antenna 2-1 to 2-N Analog signal processing unit 3-1 to 3-N A / D conversion unit 4, 4a, 4b, 4c, 4d, 4e Digital signal demodulation unit 5 Deinterleave unit 6, 6a Decoding unit 7, 9 Adder 8 Interleaving unit 10-1 to 10-N FFT unit 11 Parallel / serial conversion unit 41, 41a, 41b First signal determination unit 42 Transmission path estimation unit 43 Signal point information generation unit 44 , 44a, 44b, 44c Second signal determination unit 45 Soft determination calculation unit 411 Fourth signal determination unit 412 Fifth signal determination unit 413 Weight coefficient calculation unit 441 Symbol determination unit 442 Replica subtraction unit 443 Third signal determination Unit 444 bit determination unit 445 bit probability calculation unit 446 expected value calculation unit 447 likelihood information calculation unit

Claims (6)

A receiver that demodulates signals received by a plurality of antennas (referred to as received signals) using a maximum likelihood determination method,
First signal determination means for estimating a hard decision sequence based on the received signal and a transmission path response of the received signal, and outputting the hard decision sequence and first likelihood information corresponding to the hard decision sequence; ,
A new signal obtained by excluding signal points at which the bit value of the target bit position matches the bit value of the same bit position in the hard decision sequence from all possible signal points of the received signal for each bit position of the transmission symbol Signal point information generating means for generating a point arrangement;
Signal determination based on the received signal and the transmission path response is performed using the new signal point arrangement generated for each bit position of the transmission symbol, and second likelihood information corresponding to the signal determination result is calculated. Second signal determining means for outputting
Soft decision calculation means for calculating a soft decision value using the first likelihood information and the second likelihood information;
A receiving apparatus comprising: The second signal determination means includes
Third signal determination means for separating the received signal into a symbol including the bit for which the soft decision value is to be calculated and a symbol other than the symbol, and performing signal determination for the symbol including the bit for which the soft decision value is to be calculated;
A replica subtracting means for generating a replica using the signal determination result and the transmission path response, and generating a signal obtained by subtracting the replica from the received signal;
Using a signal generated by the replica subtraction means and the transmission path response, a fourth determination means for performing signal determination of symbols not determined by the third signal determination means among the transmission symbols;
The receiving apparatus according to claim 1, further comprising:   The second signal determination means calculates a soft decision value for a bit sequence in which a bit for which a soft decision value is to be calculated is different from the hard decision sequence, based on the received signal, the transmission path response, and the new signal point arrangement. , Using the soft decision value, calculate an expected value of a transmission symbol corresponding to a bit sequence in which the bit of the soft decision value is different from the hard decision sequence, and obtain second likelihood information based on the expected value. The receiving apparatus according to claim 1, wherein the receiving apparatus performs calculation. Likelihood information correction means for correcting the output of the second determination means using a predetermined correction coefficient;
Further comprising
The first signal determination means performs signal determination by the same technique as the second signal determination means based on the received signal and the transmission path response, and a third likelihood corresponding to the signal determination result 4. The reception according to claim 1, wherein the predetermined correction coefficient is calculated using the first likelihood information and the third likelihood information after calculating the information. apparatus. Decoding means for repeatedly executing decoding of soft input / soft output over a predetermined number of times based on the soft decision value calculated by the soft decision calculation means;
Further comprising
5. The method according to claim 1, wherein the first signal determination unit and the second signal determination unit repeatedly perform signal determination for the predetermined number of times based on an output of the decoding unit. The receiving device described in 1. A signal conversion means for converting a time domain received signal into a frequency domain signal and calculating a subcarrier signal;
Further comprising
When the received signal is a multicarrier modulation signal, the first signal determining means and the second signal determining means determine the subcarrier signal as a received signal, and the soft decision calculating means The receiving apparatus according to claim 1, wherein a soft decision value is calculated for each carrier signal.

Images