JP2013162312A - Receiving apparatus and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enhance error correction capability in a receiving apparatus for receiving an OFDM signal.SOLUTION: A receiving apparatus 1 includes an input processing unit 12 for generating a complex baseband signal from an OFDM signal, a transmission path response calculation unit 13 for calculating the transmission path response of each carrier, an equalization unit 14 for generating an estimate of a transmission signal, a noise variance calculation unit 16 for calculating the noise variance of each carrier by multiplying the band noise variance of an OFDM signal by a weighting factor for each carrier, a likelihood ratio calculation unit 18 for calculating the likelihood ratio of each bit transmitted using the carrier noise variance and the estimate of a transmission signal, a reception signal bit rearrangement unit 19 for changing the bit arrangement so that the estimate of a transmission signal corresponds with the modulation system of a transmission signal, a bit noise variance generation unit 20 for outputting the carrier noise variance as a noise variance in unit of bit, and an error correction unit 21 for decoding the estimate of a bit transmitted using the likelihood ratio, the estimate of a transmission signal after bit rearrangement, and the noise variance in unit of bit.

Description

  The present invention relates to a receiving apparatus and a program for receiving OFDM (Orthogonal Frequency Division Multiplexing).

  ISDB-T (Integrated Services Digital Broadcasting-Terrestrial), a Japanese terrestrial digital broadcasting system, realizes high-definition broadcasting (or multiple standard-definition broadcasting) for fixed receivers. In the next-generation terrestrial digital broadcasting system, it is required to provide services with a larger amount of information such as 3D high-definition broadcasting and super high-definition with 16 times the resolution of high-definition instead of conventional high-definition.

  In recent years, a MIMO (Multiple Input Multiple Output) system using a plurality of transmission / reception antennas has been proposed as a method for expanding the data transmission capacity by radio. In a transmission system using MIMO, space division multiplexing (SDM) and space time codes (STC) are performed. As an implementation example of SDM, a polarization MIMO system that uses both horizontal polarization and vertical polarization simultaneously has been proposed.

  As error correction techniques used for digital transmission, turbo codes and LDPC (Low Density Parity Check) codes have excellent error correction capabilities and are used in many transmission systems. The decoding process of these codes is performed by estimating the probability that each bit is 0 or 1 using a parameter called a log likelihood ratio (LLR) and using the calculated log likelihood ratio. Repeat. Although the log likelihood ratio is not generally updated, a method for updating the log likelihood ratio is also known (see, for example, Non-Patent Document 1). In Non-Patent Document 1, two of the quadrature amplitude modulation multiplex symbol points are determined using the n-th estimation result, and the determined two symbol points and the actual reception symbol point (the estimated value of the transmission signal). The method of calculating the square Euclidean distance and the likelihood ratio and updating the likelihood ratio is shown.

Wadayama, “A Coded Modulation Scheme Based on Low Density Parity Check Codes”, IEICE TRANS.FUNDAMENTALS, Vol.E84-A, No.10, Oct.2001

In the method of updating the log likelihood ratio shown in Non-Patent Document 1, the log likelihood ratio after update based on the error correction estimation word c (hereinafter also referred to as “update log likelihood ratio”) λ ′ is , Received symbol point coordinates (estimated value of transmission signal) x ^ and noise variance σ ^ ² are expressed by equation (1). Here, P (· | ·) is a conditional probability, and is expressed by equations (2) and (3). The noise variance σ ^ ² is a constant value in the entire band.

When data is gray-coded by mapping processing, c ₀ and c ₁ are different from each other only in the k-th bit (0 ≦ k ≦ μ−1) in one symbol data, and are expressed by Expression (4). μ is the number of bits corresponding to the modulation method. For example, in the case of 1024QAM, μ = 10.

Conventionally, in Equation (2), the noise variance σ ^ ² is calculated over the entire band, so a constant value is used. However, in MIMO transmission, the noise variance differs from carrier to carrier due to factors such as level differences between receiving antennas and multipath, and therefore an error may occur in the updated log likelihood ratio. Since obtaining the updated log likelihood ratio as accurately as possible in the decoding process affects the performance of error correction, there is a problem that the error of the updated log likelihood ratio hinders the error correction effect.

  In order to solve the above-described problem, an object of the present invention is to provide a receiving apparatus capable of improving error correction capability by utilizing C / N information in units of carriers and increasing the accuracy of an updated log likelihood ratio. To provide a program.

  In order to solve the above problems, a receiving apparatus according to the present invention is a receiving apparatus that receives an OFDM signal encoded by an error correction code that can be decoded by soft decision decoding in a SISO system, and that receives the received OFDM signal. An input processing unit that generates a complex baseband signal by Fourier transform, a transmission line response calculation unit that calculates a transmission line response of each carrier based on a known pilot signal included in the complex baseband signal, and the complex An equalization unit that divides the baseband signal by the transmission path response to generate an estimated value of the transmission signal, calculates a band noise variance of the received OFDM signal, and sets a weighting factor for each carrier with respect to the band noise variance. A noise variance calculation unit that calculates a carrier noise variance that is a noise variance of each carrier by multiplication, and uses the carrier noise variance and the estimated value of the transmission signal. A likelihood ratio calculation unit that calculates a likelihood ratio of each transmitted bit, and a received signal bit rearrangement unit that changes the bit arrangement so that the estimated value of the transmission signal corresponds to the modulation scheme of the transmission signal A bit noise variance generator for generating bit-wise noise variance from the carrier noise variance, a likelihood ratio calculated by the likelihood ratio calculator, and a bit rearrangement generated by the received signal bit rearranger An error correction decoding unit that calculates an estimated word, which is an estimated value of a transmitted bit, using an estimated value of a later transmission signal and a noise variance of a bit unit generated by the bit noise variance generating unit; It is characterized by that.

  In order to solve the above problems, a receiving apparatus according to the present invention is a receiving apparatus that receives an OFDM signal encoded by an error correction code that can be decoded by soft-decision decoding in a MIMO system. An input processing unit that generates a plurality of complex baseband signals corresponding to a plurality of receiving antennas by performing Fourier transform on the OFDM signal received from the antenna, and estimates a transmission signal by multiplying the complex baseband signal by a weight matrix A MIMO detection unit that generates a value; a MIMO detection unit that generates an estimated value of a transmission signal from the complex baseband signal using the transmission line response; and a band noise variance of the received OFDM signal; The noise component is calculated by multiplying the variance by the weighting factor for each carrier and calculating the carrier noise variance, which is the noise variance of each carrier. A calculation unit; a likelihood ratio calculation unit that calculates a likelihood ratio of each transmitted bit using the carrier noise variance and the estimated value of the transmission signal; and A received signal bit rearrangement unit that changes the bit arrangement so as to correspond to the scheme, a bit noise variance generation unit that generates bit-unit noise variance from the carrier noise variance, and a likelihood calculated by the likelihood ratio calculation unit Of the transmitted bit using the frequency ratio, the estimated value of the transmission signal after the bit rearrangement generated by the received signal bit rearrangement unit, and the noise variance of the bit unit generated by the bit noise variance generation unit. And an error correction decoding unit that decodes the estimated value.

  Furthermore, in the receiving apparatus, the noise variance calculation unit multiplies the band noise variance by a weighting factor for each carrier so that a carrier having a larger amplitude of the transmission line response or an element of the weight matrix becomes smaller. And calculating noise variance of each carrier.

  Furthermore, in the receiving apparatus, the error correction decoding unit includes a soft decision decoding unit that calculates an estimated word by repeatedly calculating using a likelihood ratio, and a bit rearrangement generated by the received signal bit rearrangement unit A likelihood ratio optimum for calculating the likelihood ratio of the estimated word calculated by the soft decision decoding unit using the estimated value of the later transmission signal and the noise variance of the bit unit generated by the bit noise variance generating unit And the soft decision decoding unit uses the likelihood ratio calculated by the likelihood ratio calculation unit as an initial value of the likelihood ratio, and each time the iterative calculation is performed, the likelihood The estimated word is calculated using the likelihood ratio calculated by the ratio optimization unit.

  The receiving apparatus further includes a bit deinterleaving unit that deinterleaves the likelihood ratio calculated by the likelihood ratio calculating unit in units of bits, and the error correction decoding unit is calculated by the soft decision decoding unit. A re-bit interleaving unit that interleaves the estimated word in bit units, and a re-bit de-interleaving unit that de-interleaves the likelihood ratio calculated by the likelihood ratio optimization unit in bit units, When the OFDM signal is interleaved in bit units, the likelihood ratio optimizing unit calculates the likelihood ratio of the estimated word that has been interleaved by the re-bit interleaving unit, and the soft decision decoding The initial value of the likelihood ratio when the OFDM signal is interleaved in bit units. The likelihood ratio calculated by the likelihood ratio calculation unit is used, and the estimated word is calculated using the likelihood ratio deinterleaved by the re-bit deinterleaving unit each time an iterative calculation is performed. It is characterized by that.

  In order to solve the above problems, a program according to the present invention causes a computer to function as the receiving device.

  ADVANTAGE OF THE INVENTION According to this invention, in the receiver which receives an OFDM signal, error correction capability can be improved by utilizing CN ratio information in a carrier unit, and improving the precision of update log likelihood ratio.

It is a block diagram which shows the structural example of the receiver which concerns on Example 1 of this invention. It is a block diagram which shows the structural example of the noise dispersion | distribution calculation part in the receiver which concerns on Example 1 of this invention. It is a block diagram which shows the structural example of the error correction decoding part in the receiver which concerns on Example 1 of this invention. It is a block diagram which shows the structural example of the receiver which concerns on Example 2 of this invention. It is a block diagram which shows the structural example of the error correction decoding part in the receiver which concerns on Example 2 of this invention. It is a block diagram which shows the structural example of the receiver which concerns on Example 3 of this invention. It is a block diagram which shows the structural example of the noise dispersion | distribution calculation part in the receiver which concerns on Example 3 of this invention. It is a block diagram which shows the structural example of the error correction decoding part in the receiver which concerns on Example 3 of this invention. It is a block diagram which shows the structural example of the receiver which concerns on Example 4 of this invention. It is a graph which shows the 1st example of the simulation result of a BER characteristic. It is a graph which shows the 2nd example of the simulation result of a BER characteristic.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the following description, the subscript i means the carrier number of the OFDM signal. The subscript i is omitted as appropriate.

  In the first embodiment, a receiving apparatus that receives an OFDM signal encoded by an error correction code that can be decoded by soft decision decoding in a SISO system will be described. An error correction code that can be decoded by soft decision decoding is an LDPC code, a turbo code, or the like. The decoding of the LDPC code and the turbo code is repeatedly performed based on the soft decision value.

  FIG. 1 is a block diagram illustrating a configuration example of a receiving apparatus according to Embodiment 1 of the present invention. As illustrated in FIG. 1, the reception device 1-1 includes a reception antenna 11, an input processing unit 12, a transmission path response calculation unit 13, an equalization unit 14, a first frequency deinterleaving unit 15, and noise dispersion. The calculator 16, the second frequency deinterleaver 17, the likelihood ratio calculator 18, the received signal bit rearranger 19, the bit noise variance generator 20, and the error correction decoder 21 are provided. In addition, when the transmission apparatus which transmits a transmission signal with respect to the receiving apparatus 1-1 does not perform a frequency interleaving process, it is not necessary to provide the 1st frequency deinterleaving part 15 and the 2nd frequency deinterleaving part 17.

  The input processing unit 12 receives a signal transmitted from the transmission device via the reception antenna 11. As shown in FIG. 1, the input processing unit 12 includes a GI (Guard interval) removal unit 121, a Fourier transform unit 122, and a pilot signal extraction unit 123.

  The GI removal unit 121 performs orthogonal demodulation processing on the OFDM signal received via the reception antenna 11 to generate a baseband signal. Then, the GI removal unit 121 extracts the effective symbol signal by removing the guard interval, and outputs it to the Fourier transform unit 122.

The Fourier transform unit 122 performs discrete Fourier transform processing on the effective symbol signal extracted by the GI removal unit 121 to generate a complex baseband signal y _i in the frequency domain, and generates a pilot signal extraction unit 123, an equalization unit 14 and the noise variance calculation unit 16.

The pilot signal extraction unit 123 extracts a known pilot signal called SP (Scattered Pilot) from the complex baseband signal y _i generated by the Fourier transform unit 122 and outputs the pilot signal to the transmission path response calculation unit 13.

The transmission line response calculation unit 13 calculates the transmission line response H _i of each carrier based on the pilot signal extracted by the pilot signal extraction unit 123 and outputs it to the equalization unit 14 and the noise variance calculation unit 16.

を生成し、尤度比算出部１８に出力する。 The equalization unit 14 performs complex division on the complex baseband signal y _i generated by the Fourier transform unit 122 by the transmission line response H _i calculated by the transmission line response calculation unit 13, removes transmission line distortion, and transmits the transmission signal. Estimated value of

Is output to the likelihood ratio calculation unit 18.

  The first frequency deinterleaving unit 15 returns the estimated value of the transmission signal generated by the equalization unit 14 to the original order by deinterleaving processing in the frequency direction (data rearranged by the frequency interleaving processing of the transmitting device). Process), and outputs the deinterleaved data to the likelihood ratio calculation unit 18.

  The receiving apparatus 1-1 needs to calculate the noise variance in order to calculate the log likelihood ratio necessary for LDPC decoding. Although the noise variance of the entire band may be calculated from the data carrier deinterleaved by the first frequency deinterleave unit 15, in order to calculate a more accurate noise variance, as described later, a carrier that is not a data carrier is used. For example, it is necessary to estimate the noise variance of the entire band using carriers called AC (Auxiliary Channel) and TMCC (Transmission and Multiplexing Configuration Control), and to weight each carrier with noise variance. Therefore, in the receiving apparatus 1-1 illustrated in FIG. 1, the noise variance calculation unit 16 is not provided between the first frequency deinterleave unit 15 and the likelihood ratio calculation unit 18 but before the first frequency deinterleave unit 15. It is arranged.

を算出し、第２周波数デインターリーブ部１７に出力する。詳細については後述する。 The noise variance calculation unit 16 performs noise variance for each carrier (hereinafter referred to as “carrier noise variance”).

Is output to the second frequency deinterleaver 17. Details will be described later.

  The second frequency deinterleaving unit 17 performs a deinterleaving process on the carrier noise variance generated by the noise variance calculating unit 16, and the deinterleaved data is converted into a likelihood ratio calculating unit 18 and a bit noise variance generating unit 20. Output to.

The likelihood ratio calculation unit 18 estimates the transmission signal x ^ _i deinterleaved by the first frequency deinterleaver 15 and the carrier noise variance σ ^ _i deinterleaved by the second frequency deinterleaver 17. ² is used to calculate the likelihood ratio λ _n of each bit of the estimated value of the transmission signal and output it to the error correction decoding unit 21.

As the likelihood ratio, a log-likelihood ratio (LLR) is generally used. log likelihood ratio lambda _n of n-th bit, the bit value b _n of n-th bit value b _n are 0 bits become likelihood function and the n-th bit of the ratio of the likelihood function becomes 1 Expressed in logarithm. That is, the log-likelihood ratio λ _n is expressed by the equations (5), (6), and (7) using the noise variance σ ² and the square Euclidean distance between the ideal symbol point s and the estimated value x ^ of the transmission signal. expressed.

を誤り訂正復号部２１に出力する。 The received signal bit rearrangement unit 19 changes the bit arrangement so that the estimated value x ^ _i of the transmission signal deinterleaved by the first frequency deinterleaving unit 15 corresponds to the modulation scheme of the transmission signal, Estimated value of transmitted signal after array

Is output to the error correction decoding unit 21.

を生成して誤り訂正復号部２１に出力する。例えば、変調方式が１０２４ＱＡＭのあるシンボルのキャリア雑音分散σ＾_ｉ ^２の値がＡであり、次のシンボルのキャリア雑音分散σ＾_ｉ ^２の値がＢである場合には、ビット単位のキャリア雑音分散σ＾_ｉｋ ^２の値は、１０ビット連続Ａであり、続いて１０ビット連続ｂとなる。 The bit noise variance generator 20 generates a carrier noise variance in bit units from the carrier noise variance σ ^ _i ² deinterleaved by the second frequency deinterleaver 17.

And output to the error correction decoding unit 21. For example, the modulation scheme is the value of the carrier noise variance sigma ^ _i ² symbols with 1024QAM is A, when the value of the carrier noise variance sigma ^ _i ² of the next symbol is B, the carrier noise bitwise The value of the variance σ ^ _ik ² is 10-bit continuous A, and subsequently 10-bit continuous b.

The error correction decoding unit 21 has a likelihood ratio λn calculated by the likelihood ratio calculation unit 18, an estimated value x ^ _i ′ of the transmission signal after bit rearrangement generated by the received signal bit rearrangement unit 19, and a bit Using the bit-wise carrier noise variance σ _ik ^ ² generated by the noise variance generator 20, the transmitted bit estimate c is decoded. Details will be described later.

[Noise variance calculation unit]
Next, the noise variance calculation unit 16 will be described in detail. FIG. 2 is a block diagram illustrating a configuration example of the noise variance calculation unit 16. As shown in FIG. 2, the noise variance calculation unit 16 includes a band noise variance calculation unit 161 and a carrier noise variance calculation unit 162.

を算出し、キャリア雑音分散算出部１６２に出力する。 Band noise variance calculator 161 is a band noise variance of the received OFDM signal (average noise variance over the entire bandwidth of the OFDM signal).

Is output to the carrier noise variance calculation unit 162.

As a method of calculating the band noise variance, a method of calculating from a MER (Modulation Error Ratio) value is known. The MER value of a carrier having a large modulation multi-value number cannot obtain a reliable value in the low CN (Carrier to Noise Ratio) region, and the noise variance calculated from the MER value is not a reliable value. Therefore, when the transmission signal is an OFDM signal in which carriers modulated by 64QAM, QPSK, BPSK, etc. are mixed as in the transmission system of terrestrial digital television broadcasting, the carrier having a small modulation multi-level number (in this case, BPSK) It is possible to obtain better accuracy by calculating the noise variance from the MER value. For example, in the case of digital terrestrial broadcasting ISDB-T, carriers called AC (Auxiliary Channel) and TMCC (Transmission and Multiplexing Configuration Control) are BPSK-modulated. It is preferable to calculate. The band noise variance σ- ² may be calculated using the estimated value x ^ _i of the transmission signal generated by the equalization unit 14, or the complex baseband signal y _i generated by the Fourier transform unit 122. You may calculate using.

The carrier noise variance calculation unit 162 multiplies the band noise variance σ− ² calculated by the band noise variance calculation unit 161 by a weight coefficient C for each carrier as shown in Expression (8), and transmits the transmission line response H _i. The carrier noise variance σ ^ _i ² is calculated (estimated) so that the carrier with the larger amplitude of becomes smaller. The weighting coefficient C is obtained by the ratio of α _i shown in equation (9) and α− shown in equation (10). E [•] represents an arithmetic average for a carrier with a determination flag T _i = 0 that is not a lost carrier. Since the reciprocal of α _i represents the reliability of the carrier, the carrier with higher reliability can reduce the noise variance.

[Error correction decoding unit]
Next, the error correction decoding unit 21 will be described in detail. FIG. 3 is a block diagram illustrating a configuration example of the error correction decoding unit 21. As shown in FIG. 3, the error correction decoding unit 21 includes a general soft decision decoding unit 210 and a likelihood ratio optimization unit 215.

The soft-decision decoding unit 210 performs repeated computation using the likelihood ratio λ _n calculated by the likelihood ratio calculation unit 18 by a known decoding algorithm (for example, sum-product decoding method), thereby transmitting the transmitted bits. An estimated word c that is an estimated value of is calculated. FIG. 3 shows a case where LDPC code decoding is performed, and the soft decision decoding unit 210 includes a row processing unit 211, a column processing unit 212, an estimated word calculation unit 213, and a parity check unit 214.

The row processing unit 211 uses the update expression of Expression (11) for all sets (m, n) that satisfy H _mn = 1 in the order of m = 1, 2,... M (= _NK ). The value α _mn is updated. The initial value of the prior value β _mn is zero. Here, sign (x) is a function represented by Expression (12), and f (x) is a function represented by Expression (13).

は、集合Ａ（ｍ）からｎを除いた集合を表す。 Here, A (m) represents a set of column indexes having 1 in the m-th row of the check matrix H. Set that is the range of product calculation and sum calculation

Represents a set obtained by removing n from the set A (m).

The column processing unit 212 updates the prior value β _mn using Equation (14) for all the pairs (m, n) that satisfy H _mn = 1 in the order of n = 1, 2 _,. To do.

は、集合Ｂ（ｎ）からｍを除いた集合を表す。 Here, B (n) represents a set of row indexes having 1 in the nth column of the check matrix H. Set that is the range of sum calculation

Represents a set obtained by removing m from the set B (n).

Estimating word calculation section 213, by the equation (15), the estimated word _{c = {c 1, ...,} c n, ..., c N} is calculated.

  The parity check unit 214 checks whether or not the estimated word c generated by the estimated word calculation unit 213 is a code word. If the estimated word c satisfies Equation (16), the parity check unit 214 decodes the estimated word c. Output as. When the estimated word c does not satisfy Expression (16), the parity check unit 214 uses the estimated word c generated by the estimated word calculation unit 213 as the likelihood ratio optimization unit 215 in order to perform an iterative process. Output to.

The likelihood ratio optimizing unit 215 generates the estimated value x ^ _i ′ of the transmission signal after the bit rearrangement generated by the received signal bit rearranging unit 19, and the bit unit noise generated by the bit noise variance generating unit 20. Using the variance σ ^ _ik ² , an updated log likelihood ratio λ _n ′ is calculated by Equation (17) for each bit of the estimated word c generated by the parity check unit 214. When the data is gray-coded by the mapping process, the likelihood ratio is calculated only for the two symbol points C ₀ and C ₁ . Each of C ₀ and C ₁ is data of one symbol, and only the value of the bit for which the likelihood ratio is calculated differs from 0 and 1, respectively, and the values of the other bits are the same.

The updated log likelihood ratio λ _n ′ calculated by the likelihood ratio optimization unit 215 is input to the row processing unit 211. That is, the soft decision decoding unit 210 uses the likelihood ratio λ _n calculated by the likelihood ratio calculation unit 18 as the initial value of the likelihood ratio, and each time iteratively calculates, the likelihood ratio optimization unit The estimated word c is calculated using the likelihood ratio λ _n ′ calculated by 215.

As described above, in the receiving apparatus 1 according to the first embodiment, the input processing unit 12 performs Fourier transform on the received OFDM signal to generate the complex baseband signal y _i , and the transmission path response calculation unit 13 performs the complex baseband signal. A transmission path response H _i of each carrier is calculated based on a known pilot signal included in the signal y _i , and the transmission signal is obtained by dividing the complex baseband signal y _i by the transmission path response H _i by the equalizing unit 14. The estimated value x ^ _{i of} is generated. Then, the noise variance calculation unit 16 calculates the band noise variance σ− ² of the received OFDM signal, and calculates the carrier noise variance σ ^ _i ² by multiplying the band noise variance by the weight coefficient C for each carrier. The likelihood ratio calculation unit 18 calculates the likelihood ratio λ _n of each transmitted bit using the carrier noise variance σ _i ² and the estimated value x ^ _i of the transmission signal. Then, the received signal bit rearrangement unit 19 changes the bit arrangement of the estimated value x ^ _i of the transmission signal so as to correspond to the modulation scheme of the transmission signal, and the bit noise variance generation unit 20 performs the carrier noise variance σ ^ The carrier noise variance σ ^ _ik ² in bit units is generated from _i ^2, and is generated by the error correction decoding unit 21 by the likelihood ratio λ _n calculated by the likelihood ratio calculation unit 18 and the received signal bit rearrangement unit 19. The estimated value x ^ _i ′ of the transmission signal after bit rearrangement and the carrier noise variance σ ^ _ik ² in bit units generated by the bit noise variance generator 20 are the estimated values of the transmitted bits. The estimated word c is calculated. As described above, since the receiving apparatus 1 in the SISO system uses the carrier noise variance σ ^ _ik ² in units of bits, it is possible to obtain a more accurate likelihood ratio and improve the error correction capability. .

The noise variance calculation unit 16 multiplies the band noise variance σ− ² by the weighting coefficient C for each carrier, and calculates the carrier noise variance σ ^ _i ² so that the carrier with the larger amplitude of the transmission path response H _i becomes smaller. Is preferred. Accordingly, the carrier noise variance can be reduced as the carrier response H _i has a larger amplitude, that is, the reliability is higher, and the carrier noise variance σ ^ _i ² can be appropriately obtained.

In more detail, the error correction decoding unit 21 calculates the estimated word c by the iterative calculation using the likelihood ratio λ _n by the soft decision decoding unit 210, and the received signal by the likelihood ratio optimization unit 215. using the estimated value x ^ _i ', and the carrier noise variance sigma ^ _ik ² bit unit generated by the bit noise variance generator 20 of the generated bit rearrangement after the transmission signal by the bit rearrangement section 19, soft The update likelihood ratio λ _n ′ of the estimated word c calculated by the determination decoding unit 210 is calculated. The soft decision decoding unit 210 uses the likelihood ratio λ _n calculated by the likelihood ratio calculation unit 18 as an initial value of the likelihood ratio, and each time the iterative calculation is performed, the likelihood ratio optimization unit 215 performs the calculation. The estimated word c is calculated using the calculated likelihood ratio λ _n ′. As a result, the accuracy of the estimated word c can be improved, the convergence speed of the decoding algorithm can be increased, and the number of iterations can be reduced.

  Next, a receiving apparatus when the transmitting apparatus performs bit interleave processing in the SISO system will be described. FIG. 4 is a block diagram illustrating a configuration example of the receiving apparatus according to the second embodiment of the present invention. As illustrated in FIG. 4, the reception device 1-2 includes a reception antenna 11, an input processing unit 12, a transmission path response calculation unit 13, an equalization unit 14, a first frequency deinterleave unit 15, and noise dispersion. Calculation unit 16, second frequency deinterleaving unit 17, likelihood ratio calculation unit 18, received signal bit rearrangement unit 19, bit noise variance generation unit 20, error correction decoding unit 21-2, bit decimation unit An interleave unit 22. The receiving device 1-2 shown in FIG. 4 further includes a bit deinterleaving unit 22 as compared with the receiving device 1-1 of the first embodiment shown in FIG. Is different. Since other configurations are the same as those in the first embodiment, the description thereof is omitted.

The bit deinterleaving unit 22 deinterleaves the likelihood ratio λ _n calculated by the likelihood ratio calculation unit 18 in bit units, and converts the data rearranged by the bit unit interleaving processing on the transmission device side to the original Return to order.

The error correction decoding unit 21-2 calculates the estimated value c of the transmitted bits using the likelihood ratio λ _n deinterleaved by the bit deinterleaving unit 22. FIG. 5 is a block diagram illustrating a configuration example of the error correction decoding unit 21-2 according to the second embodiment. As shown in FIG. 5, the error correction decoding unit 21-2 includes a row processing unit 211, a column processing unit 212, an estimated word calculation unit 213, and a soft decision decoding unit 210 having a parity check unit 214, and a likelihood. A power ratio optimizing unit 215, a rebit interleaving unit 216, and a rebit deinterleaving unit 217. The error correction decoding unit 21-2 is different from the error correction decoding unit 21 according to the first embodiment illustrated in FIG. 3 in that a rebit interleaving unit 216 and a rebit deinterleaving unit 217 are further provided. .

  The parity check unit 214 outputs the estimated word c generated by the estimated word calculation unit 213 to the re-bit interleave unit 216 when the estimated word c does not satisfy the above-described equation (16).

The likelihood ratio optimization unit 215 calculates an updated log likelihood ratio λ _n ′ using the estimated value x ^ _i of the transmission signal, the carrier noise variance σ ^ _ik ^2, and the estimated word c. The estimated value x ^ _i and the carrier noise variance σ ^ _ik ² are arranged corresponding to the bit interleaved transmission signal, and the estimated word c is arranged corresponding to the bit deinterleaved transmission signal. . In order to calculate the update likelihood ratio by the likelihood ratio optimization unit 215, it is necessary to unify the arrangement.

Therefore, re-bit interleaving section 216 performs the same bit-unit interleaving process on estimated word c input from parity check section 214 as that on the transmission apparatus side, and outputs the result to likelihood ratio optimization section 215. As a result, the likelihood ratio optimization unit 215 uses the estimated value x ^ _i , the carrier noise variance σ ^ _ik ² , and the estimated log likelihood ratio c corresponding to the transmission signal subjected to the bit interleave processing to update the log likelihood ratio. λ _n ′ can be calculated. The likelihood ratio optimization unit 215 outputs the calculated update log likelihood ratio λ _n ′ to the re-bit deinterleaving unit 217.

On the other hand, the decoding process is performed by repeatedly calculating the likelihood ratio λ _n subjected to the bit deinterleaving process by the bit deinterleaving unit 22 as an initial value. Therefore, the re-bit deinterleaving unit 217 deinterleaves the updated log likelihood ratio λ _n ′ input from the likelihood ratio optimization unit 215 again in units of bits similarly to the bit deinterleaving unit 22. Then, the updated log likelihood ratio λ _n ′ subjected to the bit deinterleaving process is output to the row processing unit 211.

As described above, the receiving apparatus 1-2 according to the second embodiment further performs bit deinterleaving on the receiving apparatus 1 to deinterleave the likelihood ratio λ _n calculated by the likelihood ratio calculating unit 18 in units of bits. The unit 22 is provided. In this case, the error correction decoding unit 21-2 uses the likelihood ratio λ _n deinterleaved by the bit deinterleave unit 22 and the estimated value of the transmission signal after the bit rearrangement generated by the received signal bit rearrangement unit 19. The estimated word c is calculated using x ^ _i and the carrier noise variance σ ^ _ik ² in bit units generated by the bit noise variance generator 20. Therefore, the error correction decoding unit 21-2 interleaves the estimated word c calculated by the soft decision decoding unit 210 by the re-bit interleaving unit 216 in bit units, and the re-bit deinterleaving unit 217 performs optimal likelihood ratio optimization. The likelihood ratio λ _n ′ calculated by the conversion unit 215 is deinterleaved in units of bits. Likelihood ratio optimization section 215 calculates the likelihood ratio of the estimated word subjected to the interleaving process by rebit interleaving section 216, and soft decision decoding section 210 uses likelihood ratio calculation section 18 as an initial value of the likelihood ratio. When the iterative calculation is performed using the likelihood ratio calculated by the above, the estimated word c is calculated using the likelihood ratio deinterleaved by the re-bit deinterleave unit 217 each time. Thereby, in the SISO system, the estimated word c can be correctly calculated even when the OFDM signal is interleaved in bit units on the transmission device side.

  Next, a case where the number of reception antennas is 2 will be described as an example for a reception apparatus in a MIMO system. FIG. 6 is a block diagram illustrating a configuration example of the receiving apparatus according to the third embodiment of the present invention. As illustrated in FIG. 6, the reception device 1-3 according to the third embodiment includes a reception antenna 11 (11a and 11b), an input processing unit 12-3 (12-3a and 12-3b), and a transmission path response calculation unit. 13-3, MIMO detector 23, first frequency deinterleaver 15-3, noise variance calculator 16-3, second frequency deinterleaver 17-3, likelihood ratio calculator 18-3 A received signal bit rearrangement unit 19-3, a bit noise variance generation unit 20-3, and an error correction decoding unit 21-3. In addition, when the transmission apparatus which transmits a transmission signal with respect to the receiver 1-3 does not perform a frequency interleaving process, it is necessary to provide the 1st frequency deinterleaving part 15-3 and the 2nd frequency deinterleaving part 17-3. There is no. The input processing unit 12-3 includes a GI removal unit 121-3 (121-3a and 121-3b), a Fourier transform unit 122-3 (122-3a and 122-3b), and a pilot signal extraction unit 123-3 ( 123-3a and 123-3b).

  The GI removal unit 121-3 generates a baseband signal by performing orthogonal demodulation processing on the OFDM signal received via the plurality of reception antennas 11. And GI removal part 121-3 removes a guard interval, extracts an effective symbol signal, and outputs it to Fourier-transform part 122-3.

The Fourier transform unit 122-3 performs discrete Fourier transform on the OFDM signals received from the plurality of reception antennas 11 to generate a frequency domain complex baseband signal corresponding to the plurality of reception antennas. That is, the Fourier transform unit 122-3a performs a discrete Fourier transform on the OFDM signal received from the receiving antenna 11a to generate a complex baseband signal y _i1 in the frequency domain, and sends it to the MIMO detection unit 23 and the noise variance calculation unit 16-3. Output. The Fourier transform unit 122-3b performs a discrete Fourier transform on the OFDM signal received from the reception antenna 11b to generate a complex baseband signal y _i2 in the frequency domain, and outputs the complex baseband signal y _i2 to the MIMO detection unit 23 and the noise variance calculation unit 16-3. .

The pilot signal extraction unit 123-3 extracts a known pilot signal (SP signal) from the complex baseband signal y _i generated by the Fourier transform unit 122-3, and outputs it to the transmission path response calculation unit 13-3.

The transmission line response calculation unit 13-3 calculates the transmission line response H _i using the pilot signal y _i extracted by the pilot signal extraction unit 123-3, and outputs the transmission line response H _i to the MIMO detection unit 23. The transmission path response H _i of 2 × 2 MIMO transmission is expressed by Expression (20).

Here, each element h _i11 , h _i12 , h _i21 , h _i22 of the transmission path response H _i is a complex number. h _i11 represents the state of the transmission path from the transmission antenna 10a (not shown) to the reception antenna 11a, h _i12 represents the state of the transmission path from the transmission antenna 10b (not shown) to the reception antenna 11a, and h _{i21 Represents} the state of the transmission path from the transmission antenna 10a to the reception antenna 11b, and _hi22 represents the state of the transmission path from the transmission antenna 10b to the reception antenna 11b.

を生成し、第１周波数デインターリーブ部１５−３及び雑音分散算出部１６−３に出力する。 MIMO detector 23, by known techniques, such as ZF (Zero Forcing) or MMSE (Minimum Mean Squared Error), by multiplying the weight matrix _{W i} into a complex baseband signal generated by the Fourier transform unit 122-3 Estimated value of transmitted signal

Is output to the first frequency deinterleaver 15-3 and the noise variance calculator 16-3.

When MIMO detection is performed using ZF, the weight matrix _Wi is calculated by Expression (21).

Further, when performing MIMO detection by MMSE, the weight matrix _Wi is calculated by the equation (22). Here, I is an Nr × Nr unit matrix where the number of receiving antennas is Nr. Further, γ is an SN ratio (Signal to Noise Ratio), and Nt is the number of transmitting antennas.

The first frequency deinterleaving unit 15-3 converts the estimated values x ^ _i1 and x ^ _i2 of the transmission signal generated by the MIMO detection unit 23 into one stream, and then performs deinterleaving processing between the frequency direction and the receiving antenna ( The data rearranged by the interleaving process between the frequency direction of the transmitting apparatus and the transmitting antenna is returned to the original order), and the deinterleaved data is output to the likelihood ratio calculation unit 18-3.

を算出し、第２周波数デインターリーブ部１７−３に出力する。詳細については後述する。 The noise variance calculation unit 16-3 performs carrier noise variance

Is output to the second frequency deinterleaver 17-3. Details will be described later.

The second frequency deinterleaving unit 17-3 converts the carrier noise variances σ ^ _i1 ² and σ ^ _i2 ² generated by the noise variance calculation unit 16-3 into one stream, and then demultiplexes them between the frequency direction and the receiving antenna. Interleaving is performed and output to the likelihood ratio calculation unit 18-3 and the bit noise variance generation unit 20-3.

Likelihood ratio calculation section 18-3 uses second frequency deinterleaving section 17-3 to demodulate transmission signal estimated values x ^ _i1 and x ^ _i2 deinterleaved by first frequency deinterleaving section 15-3. The likelihood ratio λ _n of each bit of the estimated value of the transmission signal is calculated using the carrier noise variances σ ^ _i1 ² and σ ^ _i2 ² subjected to the interleaving process, and is output to the error correction decoding unit 21-3.

を誤り訂正復号部２１−３に出力する。 The reception signal bit rearrangement unit 19-3 uses the estimated values x ^ _i1 and x ^ _i2 of the transmission signal deinterleaved by the first frequency deinterleaving unit 15-3 so as to correspond to the modulation scheme of the transmission signal. Estimated value of transmitted signal after changing the bit arrangement

Is output to the error correction decoding unit 21-3.

を生成して、誤り訂正復号部２１−３に出力する。 The bit noise variance generation unit 20-3 generates a carrier noise variance in bit units from the carrier noise variances σ ^ _i1 ² and σ ^ _i2 ² deinterleaved by the second frequency deinterleaver 17-3.

Is output to the error correction decoding unit 21-3.

The error correction decoding unit 21-3 uses the likelihood ratio λ _n generated by the likelihood ratio calculation unit 18-3, and the estimated value x ^ _i1 of the transmission signal bit rearranged by the received signal bit rearrangement unit 19-3. An estimated word that is an estimated value of the transmitted bit using ', x ^ _i2 ' and the bit-wise carrier noise variance σ ^ _i1 ² , σ ^ _i2 ² generated by the bit noise variance generator 20-3 Decrypt c.

[Noise variance calculation unit]
Next, the noise variance calculation unit 16-3 will be described in detail. FIG. 7 is a block diagram illustrating a configuration example of the noise variance calculation unit 16-3. As illustrated in FIG. 7, the noise variance calculation unit 16-3 includes band noise variance calculation units 161-3a and 161-3b and carrier noise variance calculation units 162-3a and 162-3b.

を算出し、キャリア雑音分散算出部１６２−３ａに出力する。同様に、帯域雑音分散算出部１６１−３ｂは、受信したＯＦＤＭ信号の帯域雑音分散

を算出し、キャリア雑音分散算出部１６２−３ｂに出力する。 The band noise variance calculation unit 161-3a performs band noise variance of the received OFDM signal (average noise variance over the entire band of the OFDM signal).

Is output to the carrier noise variance calculator 162-3a. Similarly, the band noise variance calculation unit 161-3b performs the band noise variance of the received OFDM signal.

Is output to the carrier noise variance calculator 162-3b.

を算出する。キャリア雑音分散σ＾_ｉ1 ^２は、ＭＩＭＯ検出部２３において複素ベースバンド信号ｙ_ｉに乗算される要素、つまりウェイト行列Ｗ_ｉの要素の値が大きいほど大きくなるように算出される。 Carrier noise variance calculating unit 162-3a multiplies relative band noise variance .sigma. ₁ ² calculated by the band noise variance calculation section 161-3A, the weight coefficient _{C i} derived from the weight matrix _{W i,} carrier noise dispersion

Is calculated. The carrier noise variance σ ^ _i1 ² is calculated so as to increase as the element multiplied by the complex baseband signal y _i in the MIMO detection unit 23, that is, the value of the element of the weight matrix W _i increases.

Specifically, the carrier noise variance calculating unit 162-3a, to the band noise variance .sigma. ₁ ² calculated by the band noise variance calculation section 161-3A, weights for each carrier as shown in equation (24) multiplied by the coefficient _{C i1,} it calculates a carrier noise variance σ _{^ i1} ^2. The weight matrix is expressed by Expression (23). When MIMO detection is performed by ZF, the weight matrix is calculated by Expression (21). When MIMO detection is performed by MMSE, the weight matrix is calculated by Expression (22). The The weighting coefficient C _i1 is obtained by the ratio of α _i shown in equation (25) and α− shown in equation (26). Here, E [•] represents an arithmetic mean. Since the reciprocal of α _i represents the reliability of the carrier, the carrier noise variance decreases as the reliability increases.

を算出する。重み係数Ｃ_ｉ２は式（２８）に示すβ_ｉと式（２９）に示すβ−との比で求まる。 Similarly, the second carrier noise variance calculating section 162-3b, to the band noise variance .sigma. ₂ ² calculated by the band noise variance calculation section 161-3B, the weighting factor for each carrier as shown in equation (27) Multiply by C _i2 and carrier noise variance

Is calculated. The weight coefficient C _i2 is obtained by the ratio of β _i shown in the equation (28) and β− shown in the equation (29).

FIG. 8 is a block diagram illustrating a configuration example of the error correction decoding unit 21-3. Compared with the error correction decoding unit 21 of the first embodiment, the error correction decoding unit 21-3 has the likelihood that the estimated values x ^ _i1 'and x ^ _i2 ' of the transmission signal from the reception signal bit rearrangement unit 19-3 are likelihoods. The point is input to the ratio optimization unit 215-3, and the bitwise carrier noise variances σ ^ _ik1 ² and σ ^ _ik2 ² are input from the bit noise variance generation unit 20-3 to the likelihood ratio optimization unit 215-3. Is different. Likelihood ratio optimization section 215-3 calculates updated log likelihood ratio λ _n ′ in the same manner as likelihood ratio optimization section 215. That is, in the above Expression (17) (18) (19 ), x ^ i with or _{x ^ i2} '' _{x ^ i1} instead of ', sigma _{^ ik} Instead sigma _{^ ik1} ² of ^2, sigma _{^ ik2} ² is used to calculate an updated log likelihood ratio λ _n ′.

As described above, in the receiving device 1-3 of the third embodiment, the input processing unit 12-3 performs Fourier transform on the received OFDM signal to generate the complex baseband signals y _i1 and y _i2, and calculates the transmission path response. The unit 13-3 calculates the channel response H _i of each carrier based on the known pilot signals included in the complex baseband signals y _i1 and y _i2 , and the MIMO detection unit 23 calculates the complex baseband signal y _i1. , Y _i2 is multiplied by a weight matrix W _i to generate transmission signal estimation values x ^ _i1 , x ^ _i2 . Then, the noise variance calculation unit 16-3 calculates the band noise variances σ- ₁ ² and σ- ₂ ² of the received OFDM signal, and the carrier noise is obtained by multiplying the band noise variance by the weighting coefficient C for each carrier. The variances σ ^ _i1 ² and σ ^ _i2 ² are calculated, and the carrier noise variances σ ^ _i1 ² , σ ^ _i2 ² and the estimated values x ^ _i1 and x ^ _i2 of the transmission signal are calculated by the likelihood ratio calculation unit 18-3. The likelihood ratio λ _n of each transmitted bit is calculated. Then, the received signal bit rearrangement unit 19-3 changes the bit arrangement of the estimated values x ^ _i1 and x ^ _i2 of the transmission signal so as to correspond to the modulation scheme of the transmission signal, and the bit noise variance generation unit 20- 3, the carrier noise variances σ ^ _ik1 ² , σ ^ _ik2 ² in bit units are generated from the carrier noise variances σ ^ _i1 ² , σ ^ _i2 ² , and the likelihood ratio calculating section 18 is generated by the error correction decoding section 21-3. Likelihood ratio λ _n calculated by -3, transmission signal estimated values x ^ _i1 ', x ^ _i2 ' generated by the received signal bit rearrangement unit 19-3, and bit noise variance generation An estimated word c, which is an estimated value of the transmitted bit, is calculated using the carrier noise variances σ ^ _ik1 ² and σ ^ _ik2 ² in bit units generated by the unit 20-3. Thus, since the receiving apparatus 1-3 in the MIMO system uses the carrier noise variances σ ^ _ik1 ² and σ ^ _ik2 ² in bit units, it is possible to obtain a more accurate likelihood ratio and improve error correction capability. To be able to.

Noise variance calculating section 16-3, band noise variance σ- 1 _^2, σ- _{2 2} multiplied by a weighting coefficient C for each carrier with respect to the carrier noise to be greater as the value of the elements of the weight matrix W _i is greater It is preferable to calculate the variances σ ^ _i1 ² and σ ^ _i2 ² . Thereby, the carrier noise variance can be reduced as the reliability is higher, and the carrier noise variances σ ^ _i1 ² and σ ^ _i2 ² can be obtained appropriately.

The error correction decoding unit 21-3 repeats the calculation using the likelihood ratio λ _n to calculate the estimated word c and the bit re-synchronization unit 19-3 generated by the received signal bit rearrangement unit 19-3. Using the estimated values x ^ _i1 ′, x ^ _i2 ′ of the transmitted signals after arrangement and the carrier noise variances σ ^ _ik1 ² , σ ^ _ik2 ² in bit units generated by the bit noise variance generator 20-3, The likelihood ratio optimizing unit 215-3 that calculates the update likelihood ratio λ _n ′ of the estimated word c calculated by the soft decision decoding unit 210 may be provided. The soft decision decoding unit 210 uses the likelihood ratio λ _n calculated by the likelihood ratio calculation unit 18-3 as an initial value of the likelihood ratio, and each time iteratively calculates, the likelihood ratio optimization unit The estimated word c is calculated using the likelihood ratio λ _n ′ calculated by 215-3. As a result, the accuracy of the estimated word c can be improved, the convergence speed of the decoding algorithm can be increased, and the number of iterations can be reduced.

  Next, a receiving apparatus when the transmitting apparatus performs bit interleaving processing in the MIMO system will be described. FIG. 9 is a block diagram illustrating a configuration example of a receiving apparatus according to Embodiment 4 of the present invention. As illustrated in FIG. 4, the reception device 1-4 includes a reception antenna 11, an input processing unit 12-3, a transmission path response calculation unit 13-3, a MIMO detection unit 23, and a first frequency deinterleaving unit 15. -3, noise variance calculation unit 16-3, second frequency deinterleave unit 17-3, likelihood ratio calculation unit 18-3, received signal bit rearrangement unit 19-3, and bit noise variance generation unit 20-3, a bit deinterleaving unit 22, and an error correction decoding unit 21-4. The receiving device 1-4 illustrated in FIG. 9 further includes a bit deinterleaving unit 22 as compared with the receiving device 1-3 illustrated in FIG. 6 according to the third embodiment. Is different. Since other configurations are the same as those in the third embodiment, the description thereof is omitted.

The bit deinterleave unit 22 performs a deinterleave process on the likelihood ratio λ _n calculated by the likelihood ratio calculation unit 18-3, and returns the data rearranged by the interleave process on the transmission device side to the original order.

The error correction decoding unit 21-4 calculates an estimated value c of the transmitted bits using the likelihood ratio λ _n deinterleaved by the bit deinterleaving unit 22. Although the configuration diagram is omitted, the error correction decoding unit 21-4 includes a rebit interleaving unit 216 and a rebit deinterleaving unit 217 as described in the second embodiment.

As described in the second embodiment, the re-bit interleaving unit 216 performs the same bit-unit interleaving processing as that on the transmission device side on the estimated word c input from the parity check unit 214, and the likelihood ratio optimization unit To 215. The re-bit deinterleaving unit 217 deinterleaves the updated log likelihood ratio λ _n ′ input from the likelihood ratio optimization unit 215 again in units of bits and outputs the result to the row processing unit 211.

As described above, the receiving apparatus 1-4 according to the fourth embodiment further performs a deinterleaving process on the likelihood ratio λ _n calculated by the likelihood ratio calculating unit 18-3 with respect to the receiving apparatus 1-3. A bit deinterleave unit 22 is provided. In this case, the error correction decoding unit 21-4 uses the likelihood ratio λ _n deinterleaved by the bit deinterleaving unit 22 and the transmission signal after the bit rearrangement generated by the received signal bit rearrangement unit 19-3. The estimated word c is calculated using the estimated values x ^ _i1 ′, x ^ _i2 ′ and the carrier noise variances σ ^ _ik1 ² , σ ^ _ik2 ² in bit units generated by the bit noise variance generator 20-3. It will be. Therefore, the error correction decoding unit 21-4 uses the rebit interleaving unit 216 to interleave the estimated word c calculated by the soft decision decoding unit 210 in bit units, and the rebit deinterleaving unit 217 performs optimal likelihood ratio optimization. The likelihood ratio calculated by the conversion unit 215 is deinterleaved in bit units. Likelihood ratio optimization section 215 calculates the likelihood ratio of the estimated word subjected to the interleaving process by rebit interleaving 216 section, and soft decision decoding section 210 uses likelihood ratio calculation section 18 as an initial value of the likelihood ratio. When the iterative calculation is performed using the likelihood ratio calculated by -3, the estimated word c is calculated using the likelihood ratio deinterleaved by the re-bit deinterleave unit 217 each time. Thereby, in the MIMO system, the estimated word c can be correctly calculated even when the OFDM signal is interleaved in bit units on the transmission device side.

  FIG. 10 shows a simulation result of the bit error rate (BER) characteristics of the conventional receiving apparatus and the receiving apparatus according to the present invention in a MIMO transmission line having a 6 dB level difference between the receiving antennas 11a and 11b. It is a graph. In FIG. 11, the delay amount is 1.2 [us], the DU ratio (Desired to Undesired signal ratio) is 6 dB, and the phase difference between the receiving antennas 11a and 11b (horizontal polarization and vertical polarization) is 180 °. 5 is a graph showing a simulation result of the BER characteristics of a conventional receiving apparatus and the receiving apparatus according to the present invention in a MIMO transmission path in which a one-wave multipath exists.

The line plotted with rhombus marks indicates the BER characteristics of the conventional receiving apparatus, and the line plotted with square marks indicates the BER characteristics of the receiving apparatus 1-3 or 1-4. Here, in the conventional receiving apparatus, the likelihood ratio calculation unit calculates a likelihood ratio using a constant value of noise variance, the error correction unit includes only the soft decision decoding unit 210, and updates the likelihood ratio. This means a receiving device that does not perform. The simulation conditions are 1024QAM for the modulation scheme, 3/4 for the coding rate, 1/8 for the GI ratio, 20 iterations for the sum-product decoding method, and other parameters comply with ISDB-T mode 3. doing. When the CN ratio (Carrier to Noise Ratio) at the point where the BER is 1 × 10 ⁻⁷ is read as the required CN, the required CN is improved by about 0.35 to 0.4 dB in any of the simulation results shown in FIGS. You can see that

  It should be noted that a computer can be suitably used to function as the above-described receiving devices 1, 1-2, 1-3, 1-4, and such a computer can be used as the receiving devices 1, 1-2, 1-3. , 1-4 are stored in the storage unit of the computer, and the program is read and executed by the CPU of the computer.

  Although the above embodiments have been described as representative examples, it will be apparent to those skilled in the art that many changes and substitutions can be made within the spirit and scope of the invention. Therefore, the present invention should not be construed as being limited by the above-described embodiments, and various modifications and changes can be made without departing from the scope of the claims. For example, the likelihood ratio calculation units 18 and 18-3 calculate the log likelihood ratio in the above-described embodiment, but may calculate the likelihood ratio.

  Thus, the present invention is useful for any application that receives OFDM signals.

1, 1-2, 1-3, 1-4 receiving device 11, 11a, 11b receiving antenna 12, 12-3a, 12-3b input processing unit 13, 13-3 transmission line response calculating unit 14, equalizing unit 15, 15-3 First frequency deinterleave unit 16, 16-3 Noise variance calculation unit 17, 17-3 Second frequency deinterleave unit 18, 18-3 Likelihood ratio calculation unit 19, 19-3 Received signal bit rearrangement unit 20, 20-3 bit noise variance generation unit 21, 21-2, 21-3, 21-4 error correction decoding unit 23 MIMO detection unit 121, 121-3a, 121-3b GI removal unit 122, 122-3a, 122 -3b Fourier transform unit 123, 123-3a, 123-3b Pilot signal extraction unit 161, 161-3a, 161-3b Band noise variance calculation unit 162, 162-3a, 162 -3b Carrier noise variance calculation unit 210 Soft decision decoding unit 211 Row processing unit 212 Column processing unit 213 Estimated word calculation unit 214 Parity check unit 215, 215-3 Likelihood ratio optimization unit 216 Rebit interleave unit 217 Rebit deinterleave Part

Claims (9)

In a SISO system, a receiving apparatus that receives an OFDM signal encoded by an error correction code that can be decoded by soft decision decoding,
An input processing unit that generates a complex baseband signal by performing a Fourier transform on the received OFDM signal;
A transmission line response calculation unit for calculating a transmission line response of each carrier based on a known pilot signal included in the complex baseband signal;
An equalization unit that divides the complex baseband signal by the transmission path response to generate an estimated value of the transmission signal;
A noise variance calculation unit that calculates a band noise variance of the received OFDM signal, multiplies the band noise variance by a weighting factor for each carrier, and calculates a carrier noise variance that is a noise variance of each carrier;
A likelihood ratio calculation unit that calculates a likelihood ratio of each transmitted bit using the carrier noise variance and the estimated value of the transmission signal;
The received signal bit rearrangement unit that changes the bit arrangement so that the estimated value of the transmission signal corresponds to the modulation scheme of the transmission signal;
A bit noise variance generator for generating bit-wise noise variance from the carrier noise variance;
The likelihood ratio calculated by the likelihood ratio calculation unit, the estimated value of the transmission signal after the bit rearrangement generated by the received signal bit rearrangement unit, and the bit unit generated by the bit noise variance generation unit An error correction decoding unit that calculates an estimated word that is an estimated value of a transmitted bit using noise variance;
A receiving apparatus comprising:   The noise variance calculation unit multiplies the band noise variance by a weighting factor for each carrier, and calculates the carrier noise variance so that the carrier having a larger amplitude of the transmission path response becomes smaller. Item 4. The receiving device according to Item 1. The error correction decoding unit is a soft decision decoding unit that calculates an estimated word by iteratively calculating using a likelihood ratio;
Calculated by the soft decision decoding unit using the estimated value of the transmission signal after the bit rearrangement generated by the received signal bit rearrangement unit and the noise variance of the bit unit generated by the bit noise variance generation unit. A likelihood ratio optimization unit for calculating a likelihood ratio of the estimated words,
With
The soft decision decoding unit uses the likelihood ratio calculated by the likelihood ratio calculation unit as an initial value of the likelihood ratio, and is calculated by the likelihood ratio optimization unit each time iterative calculation is performed. The receiving apparatus according to claim 1, wherein the estimated word is calculated using a likelihood ratio. A bit deinterleaving unit that deinterleaves the likelihood ratio calculated by the likelihood ratio calculation unit in units of bits;
The error correction decoding unit includes a re-bit interleaving unit that interleaves the estimated word calculated by the soft decision decoding unit in bit units and a likelihood ratio calculated by the likelihood ratio optimization unit in bit units. A re-bit deinterleaving unit for interleaving, and
When the OFDM signal is interleaved in bit units, the likelihood ratio optimizing unit calculates the likelihood ratio of the estimated word that has been interleaved by the re-bit interleaving unit, and the soft decision decoding The unit uses the likelihood ratio calculated by the likelihood ratio calculation unit as an initial value of the likelihood ratio when the OFDM signal is interleaved in units of bits, and when performing an iterative calculation, The receiving apparatus according to claim 3, wherein the estimation word is calculated using the likelihood ratio deinterleaved by the re-bit deinterleaving unit each time. In a MIMO system, a receiving apparatus that receives an OFDM signal encoded by an error correction code that can be decoded by soft decision decoding,
An input processing unit that Fourier-transforms OFDM signals received from a plurality of receiving antennas to generate a plurality of complex baseband signals corresponding to the plurality of receiving antennas;
A transmission line response calculation unit for calculating a transmission line response of each carrier based on a known pilot signal included in the complex baseband signal;
A MIMO detection unit that generates an estimated value of a transmission signal by multiplying the complex baseband signal by a weight matrix;
A noise variance calculation unit that calculates a band noise variance of the received OFDM signal, multiplies the band noise variance by a weighting factor for each carrier, and calculates a carrier noise variance that is a noise variance of each carrier;
A likelihood ratio calculation unit that calculates a likelihood ratio of each transmitted bit using the carrier noise variance and the estimated value of the transmission signal;
The received signal bit rearrangement unit that changes the bit arrangement so that the estimated value of the transmission signal corresponds to the modulation scheme of the transmission signal;
A bit noise variance generator for generating bit-wise noise variance from the carrier noise variance;
The likelihood ratio calculated by the likelihood ratio calculation unit, the estimated value of the transmission signal after the bit rearrangement generated by the received signal bit rearrangement unit, and the bit unit generated by the bit noise variance generation unit An error correction decoding unit that decodes an estimate of the transmitted bits using noise variance;
A receiving apparatus comprising:   The noise variance calculating unit multiplies the band noise variance by a weighting factor for each carrier, and calculates the noise variance of each carrier so as to increase as the value of the element of the weight matrix increases. The receiving device according to claim 5. The error correction decoding unit is a soft decision decoding unit that calculates an estimated word by iteratively calculating using a likelihood ratio;
Calculated by the soft decision decoding unit using the estimated value of the transmission signal after the bit rearrangement generated by the received signal bit rearrangement unit and the noise variance of the bit unit generated by the bit noise variance generation unit. A likelihood ratio optimization unit for calculating a likelihood ratio of the estimated words,
With
The soft decision decoding unit uses the likelihood ratio calculated by the likelihood ratio calculation unit as an initial value of the likelihood ratio, and is calculated by the likelihood ratio optimization unit each time iterative calculation is performed. The receiving apparatus according to claim 5, wherein the estimated word is calculated using a likelihood ratio. A bit deinterleaving unit that deinterleaves the likelihood ratio calculated by the likelihood ratio calculation unit in units of bits;
The error correction decoding unit includes a re-bit interleaving unit that interleaves the estimated word calculated by the soft decision decoding unit in bit units and a likelihood ratio calculated by the likelihood ratio optimization unit in bit units. A re-bit deinterleaving unit for interleaving, and
When the OFDM signal is interleaved in bit units, the likelihood ratio optimizing unit calculates the likelihood ratio of the estimated word that has been interleaved by the re-bit interleaving unit, and the soft decision decoding The unit uses the likelihood ratio calculated by the likelihood ratio calculation unit as an initial value of the likelihood ratio when the OFDM signal is interleaved in units of bits, and when performing an iterative calculation, The receiving apparatus according to claim 7, wherein the estimation word is calculated using the likelihood ratio deinterleaved by the re-bit deinterleaving unit each time. The program for functioning a computer as a receiver as described in any one of Claim 1 to 8.

Images