JP2017034338A - Receiver and interference estimation method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect an interference wave superimposed on a desired wave, regardless of bandwidth or the number of interference wave.SOLUTION: A receiver 200 receives a signal in which a desired wave transmitted by using a multi-carrier multiplexing transmission system and one or more interference wave which interferes the desired wave are superimposed. An equalizer circuit 263, a hard decision circuit 264, a mapping circuit 266, and a transmission path weighting circuit 267 generates a replica signal of a reception signal of the desired wave using a transmission signal estimated from a reception signal. An unwanted signal power calculation circuit 268 calculates power of an unwanted wave for each subcarrier by subtracting the replica signal from the reception signal. A sub-channelizing circuit 269 determines a sub channel to which a subcarrier belongs. A logarithmic likelihood expected value calculation circuit 270 and a maximizing circuit 271 estimate an interference probability which is a probability of interference wave existence on a desired wave, and interference power of the interference wave, in subcarrier basis, so as to maximize an expected value of likelihood on undesired wave power in each sub channel.SELECTED DRAWING: Figure 3

Description

  The present invention relates to a receiving apparatus and an interference estimation method.

  In recent years, the depletion of frequency resources has become a problem due to the widespread use of various wireless communication systems, and studies on superposition transmission techniques that improve frequency utilization efficiency by sharing frequencies with a plurality of wireless signals are being promoted. For example, FIG. 11 is a conceptual diagram showing the entire two wireless LAN (Local Area Network) systems having different frequency channels as an example of a combination of wireless communication systems sharing a frequency band. In the figure, the wireless communication system includes a wireless LAN base station 10a and a wireless LAN base station 10b, and a receiving device 20a composed of a personal computer or the like. The wireless LAN base station 10a communicates using the frequency band of the channel CH1 that is the center frequency fa. On the other hand, the wireless LAN base station 10b communicates using the frequency band of the channel CH5 having the center frequency fb (fa <fb).

  The receiving device 20a is arranged at a position where wireless signals of both the wireless LAN base station 10a and the wireless LAN base station 10b reach. The receiving device 20a receives a signal in which two radio signals, that is, a radio signal having a center frequency fa that is a desired signal D and a radio signal having a center frequency fb that is an interference signal U partially interfere with each other. In addition, as another example of sharing the frequency band, there is a combination of a wireless LAN system, Bluetooth (registered trademark), and WiMAX (registered trademark), and systems with different communication methods may also share frequencies. .

  Although FIG. 11 shows an example in which the continuous frequency spectrum of the upper end frequency of the desired signal D and the lower end frequency of the interference signal U is superimposed, a case where the interference signal is a relatively narrow band signal is also assumed. Also, as shown in FIG. 12, the interference signal U for the desired signal D may be composed of a plurality of narrowband signals by a transmission method such as frequency hopping. Alternatively, as shown in FIG. 13, there may be a case where interference signals U1 and U2 of different communication schemes arrive from a plurality of interference sources.

  In general, when such an interference wave exists, communication characteristics are significantly deteriorated. Therefore, on the assumption that the transmission method of the desired wave is multi-carrier and error correction coding is performed, the received signal is FEC (forward error correction) decoded while suppressing the influence of interference, There is a technique for realizing accurate transmission (see, for example, Non-Patent Document 1). Specifically, the influence of the interference wave is suppressed on the received signal before the demodulation of the desired wave. In order to suppress the influence of the interference wave, a frequency component in which the interference wave exists in the received signal is subjected to a filtering process in an RF (Radio Frequency) stage or an IF (Intermediate Frequency) stage. Alternatively, in order to suppress the influence of the interference wave, likelihood weighting processing is performed on the corresponding frequency component in the baseband region. After suppressing the influence of the interference wave in this way, the received signal is demodulated and decoded.

  In order to perform the reception process as in Non-Patent Document 1, it is necessary to detect the frequency band (interference probability of each subcarrier) in which an interference wave exists and the desired wave-to-interference wave power ratio, but these are unnecessary. A simple receiving method has also been proposed (see, for example, Patent Document 1). In this reception method, a reception replica signal of a desired signal is generated for a packet in which no error has occurred in means such as CRC (Cyclic Redundancy Check) detection as a result of provisional demodulation decoding of the reception signal. Then, a subcarrier whose residual signal power, which is a result of subtracting the received replica signal from the received signal, exceeds the threshold is identified as an interference band, and the residual signal power in the interference band is estimated as interference power.

  However, since a very large number of packets need to be transmitted until a sufficient number of packets without errors are obtained in CRC detection, the delay time until interference detection becomes a big problem. Although the interference detection is possible even if the CRC check is not passed if the equal power modulation method is used, there is still a problem that a plurality of packets are required.

  On the other hand, using the statistical property that the amplitude distribution of the noise signal and interference signal follows a Gaussian distribution, the time required for interference detection by applying an EM (Expectation Maximization) algorithm (see, for example, Non-Patent Document 2) is reduced. There is a technique for shortening (for example, see Non-Patent Document 3). In this technique, the residual signal is calculated by “subtracting the hard decision replica of the desired wave from the desired signal”. Then, L residual signals having the same number as the subcarriers of observation data are prepared after averaging k symbols, and the EM algorithm is repeatedly applied. In the EM algorithm, “E step: deriving a conditional expected value (Q function) of log likelihood for observed data” and “M step: deriving an interference parameter that maximizes the Q function”. By repeatedly applying the EM algorithm, the final interference parameters are derived. The interference parameters are interference probability and noise / interference power.

The technology of Non-Patent Document 3 will be outlined. When the set S _if (t) of subcarrier numbers belonging to the interference band at time t is known, log likelihood for the mth bit c (t, l, m) of the transmission signal at time t and subcarrier number l The ratio LLR ^Perf [c (t, l, m)] is expressed by the following equation (A) using the noise power σ _n ² and the interference power σ _if ² .

However, d _x (t, l) = | y (t, l) −x (t, l) · h (t, l) | ² , y (t, l), h (t, l), x (t, l) is a received signal, a transmission path response, and a transmitted signal at time t and subcarrier number l, respectively. X ₁ (m) and X ₀ (m) are each set to 1 in the m-th bit in modulation mapping such as PSK (Phase Shift Keying) and QAM (Quadrature Amplitude Modulation). And a set of signal points that are zero.

However, as described above, it is necessary to estimate a set S _if (t) of subcarrier numbers belonging to the interference band. Therefore, p _if (l), which is the probability that the subcarrier l receives interference, is derived, and the mixture ratio of each LLR at the time of interference and non-interference of the above-described log likelihood ratio LLR ^Perf , as shown in Expression (B). The basic idea is to obtain a mixed log-likelihood ratio LLR ^Weigh . LLR ^Weight [c (t, l, m)] is a mixed log likelihood ratio for the mth bit c (t, l, m) of the transmission signal at time t and subcarrier number l.

Japanese Patent No. 5622955

Satoshi Masuno, 1 other person, “Improvement of frequency utilization efficiency by multi-carrier superposition transmission”, IEICE Technical Report, IEICE, 2008, RCS2008-67, vol. 108, no. 188, p.85 -90 Moon, T.K, “The Expectation-Maximization Algorithm”, IEEE, Signal Processing Magazine, 1996, Vol. 13, Issue 6, p.47-60 Naotoken Yoda, 3 others, "Interference suppression by EM algorithm in OFDM transmission", IEICE, IEICE, 2014, vol. 113, no.456, p.443-447

  In Non-Patent Document 3, it is assumed that the number of interference waves is 1. However, as described above, a plurality of interference waves may actually arrive. When the wireless transmission path has the characteristics of frequency selective fading due to the influence of delayed waves, even if it is a single interference source, the incoming interference power varies greatly depending on the frequency band, and the amplitude distribution behaves differently from the Gaussian distribution. It is assumed that

In order to cope with a plurality of interference waves and interference waves that have undergone frequency selective fading, it is conceivable to implement the technique of Non-Patent Document 3 on a subcarrier basis of a desired wave. However, Non-Patent Document 3 is characterized in that observation data is prepared in the same direction as the number of subcarriers in the subcarrier direction on the assumption that the interference probability and interference power are the same for all subcarriers. Therefore, when it is simply performed in units of subcarriers, there is a problem that updating of interference parameters due to repetition of the EM algorithm does not function.
In addition, the interference parameters obtained by Non-Patent Document 3 are the interference probability and the interference power, and the noise power required for calculating the log likelihood ratio needs to be known, and the measurement method is not mentioned.

  In view of the above circumstances, an object of the present invention is to provide a receiver and an interference estimation method that can detect an interference wave superimposed on a desired wave regardless of the band and number of interference waves.

  One embodiment of the present invention is a receiving device that receives a signal in which a desired wave transmitted using a multicarrier superimposed transmission scheme and one or more interference waves that interfere with the desired wave are superimposed. A replica signal generation unit that generates a replica signal of the received signal of the desired wave using a transmission signal estimated from the signal, and unnecessary calculation of the power of the unnecessary wave for each subcarrier by subtracting the replica signal from the received signal A signal power calculation unit, a subchannelization unit that determines a subchannel to which the subcarrier belongs, and interference in the desired wave so that an expected value of likelihood for the power of the unnecessary wave of each subchannel is maximized. And an estimation unit that estimates an interference probability, which is a probability that a wave exists, and interference power of the interference wave for each subcarrier.

  One aspect of the present invention is the above-described receiving device, wherein the estimation unit averages the interference probability and the interference power estimated for the subcarriers belonging to the subchannel to obtain an interference probability for each subchannel and Estimate interference power.

  One aspect of the present invention is the above-described receiving apparatus, wherein noise power common to all subcarriers is calculated based on a known signal included in the received signal, and the interference probability is obtained for each subchannel. After that, a noise power estimation unit is further provided that updates noise power common to all subcarriers based on noise power of the subcarriers belonging to the subchannel whose average interference probability is below a predetermined threshold.

  One aspect of the present invention is the above-described receiving device, wherein the subchannelization unit determines a subchannel to which the subcarrier belongs according to the magnitude of the power of the subcarrier in the unnecessary wave.

  One aspect of the present invention is the above-described receiving apparatus, wherein a mixed log likelihood ratio calculation that calculates a mixed log likelihood ratio from the received signal using the interference probability and the interference power estimated by the estimation unit. The unit is further provided.

  One aspect of the present invention is the above-described receiving device, in which the replica signal generation unit updates the mixed log likelihood ratio with the interference probability and the interference power estimated by the estimation unit, and performs error correction decoding. A process of generating a new replica signal from the result of performing, a process of updating the power of the unnecessary wave by subtracting the new replica signal from the received signal by the unnecessary signal power calculation unit, A process in which a channelizing unit determines a subchannel to which the subcarrier in the unnecessary wave belongs; and a process in which the estimation unit estimates the interference probability and the interference power based on the updated power of the unnecessary wave; Repeat.

  One aspect of the present invention is the above-described receiving device, wherein the replica signal is a signal obtained by making a hard decision on a candidate of a transmission signal point closest to the reception signal point of the reception signal, and a new The replica signal is a signal obtained by performing error correction coding on the result of performing error correction decoding on the updated mixed log likelihood ratio, or the replica signal is an interference probability of an initial value for each subcarrier And a soft decision replica signal based on a mixed log likelihood ratio calculated from the received signal using interference power, and the new replica signal is obtained by performing a soft decision on the updated mixed log likelihood ratio It is a signal generated from the result.

  One embodiment of the present invention is a reception method performed by a reception device that receives a signal in which a desired wave transmitted using a multicarrier superimposed transmission scheme and one or more interference waves that interfere with the desired wave are superimposed. A replica signal generating step of generating a replica signal of the received signal of the desired wave using a transmission signal estimated from the received signal, and subtracting the replica signal from the received signal to subtract the power of the unnecessary wave from the subcarrier Unnecessary signal power calculation step calculated every time, subchannelization step for determining a subchannel to which the subcarrier belongs, and an expected value of likelihood for the power of the unnecessary wave of each of the subchannels, so as to maximize An estimation step for estimating an interference probability which is a probability that an interference wave exists in the desired wave and an interference power of the interference wave for each subcarrier. Is an interference estimation method with.

  According to the present invention, an interference wave superimposed on a desired wave can be detected regardless of the band and number of interference waves.

1 is a configuration diagram of a radio communication system according to a first embodiment of the present invention. It is a functional block diagram which shows the structure of the transmitter by the embodiment. It is a functional block diagram which shows the structure of the receiver by the embodiment. It is a flowchart which shows the receiving operation of the receiver by the embodiment. It is a figure which shows the example of subchannelization by the embodiment. It is a flowchart which shows the inner loop process for every subchannel of the reception process part by the embodiment. It is a functional block diagram which shows the other structure of the receiver by the embodiment. It is a functional block diagram which shows the structure of the receiver to which a prior art is applied. It is a functional block diagram which shows the structure of the receiver by 2nd Embodiment. It is a figure which shows the experimental result which simulated the detection success probability of the receiver of 1st Embodiment, and the receiver to which a prior art is applied. FIG. 2 is a conceptual diagram illustrating the entire two wireless LAN systems having different frequency channels as an example of a combination of wireless communication systems sharing a frequency band. It is a figure which shows the example in which an interference signal is comprised with a some narrowband signal. It is a figure which shows the example from which a different type of interference signal arrives from several interference sources.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

[First Embodiment]
FIG. 1 is a configuration diagram of a wireless communication system 1 according to a first embodiment of the present invention. As shown in the figure, the wireless communication system 1 includes a transmitter 100 (transmitting device) and a receiver 200 (receiving device) that communicate wirelessly. The wireless communication system 1 performs wireless communication by a multicarrier superimposed transmission method.

  FIG. 2 is a functional block diagram showing the configuration of the transmitter 100, and only functional blocks related to the present embodiment are extracted and shown. The transmitter 100 is based on OFDM (Orthogonal Frequency Division Multiplexing) transmission. As shown in the figure, the transmitter 100 includes an error correction coding circuit 110, a mapping circuit 120, an S / P (serial parallel) conversion circuit 130, an IFFT (Inverse Fast Fourier Transform) circuit 140, A click prefix adding circuit 150, a frequency converter 160, and an antenna 170 are provided.

  The error correction encoding circuit 110 performs error correction encoding on the transmission bit sequence. The mapping circuit 120 generates a transmission symbol sequence on the IQ plane from a transmission bit sequence subjected to error correction coding by PSK (Phase Shift Keying), QAM (Quadrature Amplitude Modulation), or the like. . The S / P conversion circuit 130 converts the transmission symbol string of the serial signal generated by the mapping circuit 120 into a parallel signal. The IFFT circuit 140 performs IFFT on the transmission symbol sequence converted into the parallel signal, and converts it into a time domain signal. The cyclic prefix adding circuit 150 copies the end of the block of the signal converted into the time domain to the head, and forms a transmission signal that suppresses OFDM inter-block interference due to the delayed wave. The frequency converter 160 up-converts the transmission signal generated by the cyclic prefix adding circuit 150 to a radio frequency. The antenna 170 transmits the up-converted transmission signal by radio.

  Note that an interleaver may be provided before or after the mapping circuit 120. Since the error-correction coded bits are arranged in a wide band, the frequency diversity effect can be enhanced and the tolerance to interference or frequency selective fading can be enhanced. In that case, a deinterleaver is provided at a corresponding position of the receiver 200.

  FIG. 3 is a functional block diagram showing a configuration of the receiver 200, and only functional blocks related to the present embodiment are extracted and shown. In the figure, a receiver 200 includes an antenna 210, a frequency converter 220, a cyclic prefix removal circuit 230, an FFT (Fast Fourier Transform) circuit 240, a P / S (parallel serial) conversion circuit 250, and A decoding processing unit 260 is provided.

  The antenna 210 receives a radio signal. The frequency converter 220 down-converts the radio signal received by the antenna 210 and converts it into a received signal. The cyclic prefix removal circuit 230 removes the cyclic prefix from the down-converted received signal. The FFT circuit 240 converts the received signal from which the cyclic prefix is removed into a frequency domain signal, and obtains a detection signal for each subcarrier. The P / S conversion circuit 250 rearranges the detection signals for each subcarrier obtained by the FFT circuit 240 in the time direction, and converts the parallel signals into serial signals.

  The decoding processing unit 260 includes a received signal buffer 261, a transmission path estimation circuit 262, an equalization circuit 263, a hard decision circuit 264, a switching circuit 265, a mapping circuit 266, a transmission path weight circuit 267, an unnecessary signal power calculation circuit 268, a subchannel. 269, log likelihood expectation value calculation circuit 270, maximization circuit 271, averaging circuit 272, noise power estimation circuit 273, extraction / averaging circuit 274, mixed log likelihood ratio calculation circuit 275, decoding circuit 276, and And an error correction encoding circuit 277.

The reception signal buffer 261 stores the reception signal output from the P / S conversion circuit 250. The reception signal buffer 261 is provided to absorb a processing delay from when the equalization circuit 263 performs processing until the unnecessary signal power calculation circuit 268 starts processing via the transmission line weight circuit 267.
The transmission path estimation circuit 262 extracts known signals such as preamble symbols and pilot subcarriers from the reception signal output from the P / S conversion circuit 250. The transmission path estimation circuit 262 estimates a channel response generated by radio transmission or the like from the extracted known signal for each subcarrier, and uses the equalization circuit 263 and the transmission path estimation value calculated from the estimated channel response for each subcarrier. It outputs to the road weight circuit 267.
The equalization circuit 263 equalizes the received signal for each subcarrier based on the transmission path estimation value, and outputs the equalized signal to the hard decision circuit 264 and the mixed log likelihood ratio calculation circuit 275.
The hard decision circuit 264 performs a hard decision on the transmission signal point candidate closest to the reception signal point on the IQ plane equalized by the equalization circuit 263, and estimates the transmitted bit sequence for each subcarrier. . The hard decision circuit 264 outputs the estimated bit sequence to the switching circuit 265.

  The switching circuit 265 switches the bit series output to the mapping circuit 266. The switching circuit 265 outputs the bit sequence obtained as a result of the hard decision by the hard decision circuit 264 to the mapping circuit 266 for the first time. From the second time onward, the switching circuit 265 outputs the bit series that has been subjected to error correction coding by the error correction coding circuit 277 to the mapping circuit 266.

The mapping circuit 266 modulates the input bit sequence by a modulation method (for example, PSK or QAM) using the same parameters as the transmitter 100, and generates a transmission replica signal that is a replica of the transmission signal. The mapping circuit 266 outputs the generated transmission replica signal to the transmission path weight circuit 267.
The transmission path weight circuit 267 multiplies the transmission replica signal generated by the mapping circuit 266 by the transmission path weight using the transmission path estimation value input from the transmission path estimation circuit 262, and receives the received signal of the desired wave replica for each subcarrier. A received replica signal is obtained. The transmission line weight circuit 267 outputs the obtained reception replica signal to the unnecessary signal power calculation circuit 268.

  The unnecessary signal power calculation circuit 268 reads the reception signal that is the output of the P / S conversion circuit 250 from the reception signal buffer 261. The unnecessary signal power calculation circuit 268 subtracts the reception replica signal generated by the transmission path weight circuit 267 from the read reception signal, and calculates the power for each symbol of the unnecessary wave (hereinafter referred to as “unnecessary signal power”). calculate. Furthermore, the unnecessary signal power calculation circuit 268 calculates the square value of the unnecessary signal power for each calculated symbol, and outputs it to the subchannelization circuit 269.

  The subchannelization circuit 269 performs subcarrierization that determines the subchannel to which each subcarrier belongs. For example, the subchannelization circuit 269 performs subchannelization based on the magnitude of unnecessary signal power of each subcarrier. One subchannel includes one or more subcarriers. Note that subcarrier numbers included in one subchannel are not necessarily adjacent to each other.

  Here, the symbol y (t, l) of the received signal at time t and subcarrier number l is expressed by the following equation (1).

y (t, l) = h (t, l) s (t, l) + i (t, l) + n (t, l) (1)

  In equation (1), h (t, l), s (t, l), i (t, l), and n (t, l) are the propagation path characteristics and desired wave at time t and subcarrier number l, respectively. A transmission signal, an interference signal (a signal that interferes with a desired wave), and a noise signal. The unnecessary signal power is the sum of the interference signal and the noise signal. Therefore, the square value I (t, l) of unnecessary signal power (residual power) at time t and subcarrier number l is expressed by the following equation (2).

I (t, l) = | y (t, l) −h (t, l) s (t, l) | ² (2)

The average unnecessary signal power w _l of the subcarrier number l for k symbols from the time t = t ′ is expressed by the following equation (3). However, X is a set of transmission signal candidates x.

Noise signals and interference signals follow a complex Gaussian distribution. It is assumed that the noise signal power variance is σ _n ² and the interference signal power variance is σ _if ² . Under the assumption that the k symbol period interference signal is unchanged, I (t, l) is averaged for k symbols from time t = t ′. When I (t, l) is averaged by k symbols, the noise signal and the interference signal are independent random variables that follow a normal distribution. Therefore, due to statistical properties, the average value w _l of I (t, l) is According to the mixed gamma distribution of the following equation (4). α is a ratio L _if / L of the number of interference subcarriers L _if and the total number of subcarriers L. The noise signal includes a noise signal n (t, l) for a subcarrier without an interference wave, and a noise signal n (t, l) and an interference signal i (t, l) for a subcarrier with an interference wave. It is.

However, f _G is a probability density function of a gamma distribution defined by w> 0, and is expressed by the following equation (5) using the shape parameter a and the scale parameter b. Γ (a) is a gamma function. Note that a = k.

In the following, the above w _l, is used as the "observation data" in the EM algorithm.

The decoding processing unit 260 estimates an interference parameter by calculating a Q function representing a conditional expected value of the log likelihood ratio and maximizing the Q function. The interference parameters are an interference probability that is a probability that an interference wave is included in the desired wave, and an interference power of the interference wave. Decoding processing section 260 receives the LLR calculated using the estimated interference parameter and decodes the received signal. Decoding processing section 260 creates a transmission replica signal from the estimated value of the transmission signal obtained by decoding, and further creates a reception replica signal from the transmission replica signal. The decoding processing unit 260 updates _wl with the generated received replica signal, and again estimates the interference parameter. It is desirable to repeat the estimation and decoding until the interference parameters converge. In the following, (·) ^(κ) represents parameters obtained in the κ-th iteration.

The log likelihood expectation value calculation circuit 270 calculates the log likelihood expectation value (Q function) of the following equation (6) for each subcarrier l. The interference parameter θ (l) = (β (l), σ _if ² (l)). β (l) is the estimated interference probability of subcarrier l, and σ _if ^{2 (κ)} (l) is the interference power in subcarrier l.

However, the interference probability p _{an if} the observed data w _l is the probability that belong to the gamma distribution for the interference wave, satisfies the following equation (7).

The maximization circuit 271 obtains an interference parameter θ (l) that maximizes the Q function calculated by the log likelihood expectation value calculation circuit 270. The following equation (8) is obtained under the condition that the Q function is differentiated by the interference parameter θ (l) ^(κ) = (β (l) ^(κ) , σ _if ^{2 (κ)} (l)) and becomes zero. .

The maximization circuit 271 calculates the interference probability β (l) ^(κ) and the interference power σ _if ^{2 (κ)} (l) for the subcarrier l by the equation (8). Derivation of σ _n ² will be described separately.

  At this time, if the maximum number of interference waves of a specific intensity is 1 for each subchannel, the interference parameter θ can be estimated with high accuracy by repeating the EM algorithm. Therefore, in the subchannelization processing of the subchannelization circuit 269, it is desirable for the purpose of operation to appropriately subchannel according to the magnitude of unnecessary signal power.

  As described in the present embodiment, the EM algorithm estimates points of the probability model parameters, but considers it as a random variable, a variational Bayes method for learning the posterior probability distribution, and a Monte Carlo that samples the probability distribution. An estimation method such as sampling may be used instead.

The maximization circuit 271 calculates the interference parameter θ (l) ^(κ) by performing the EM algorithm repetition processing as described above. The maximization circuit 271 sequentially outputs the interference parameter θ (l) ^(κ) when the change converges to the averaging circuit 272 as the interference parameter θ (l).

The averaging circuit 272 receives the interference parameter θ (l) = (β (l), σ _if ² (l)) from the maximizing circuit 271. For each subchannel, the averaging circuit 272 calculates the average value of the interference probability β (l) and the interference power σ _if ² (l) of the subcarriers belonging to that channel, and calculates the interference probability α (c) and the interference power. γ (c) is calculated. The averaging circuit 272 outputs the interference probability α (c) of each subchannel to the extraction / averaging circuit 274, and mixes the interference probability α (c) and interference power γ (c) of each subchannel with a mixed log likelihood ratio. The calculation circuit 275 is notified.

  The noise power estimation circuit 273 receives the reception signal output from the P / S conversion circuit 250. The noise power estimation circuit 273 estimates the noise power included in the received signal by any means. As a well-known noise power estimation method, there is a method of identifying a variance value of two or more consecutive preamble symbols arranged at the head of a radio frame as noise power, but this is not restrictive. The noise power estimation circuit 273 estimates noise power for each subcarrier, but may be averaged in the time direction.

The extraction / averaging circuit 274 receives information on the interference probability α (c) for each subcarrier from the averaging circuit 272. However, since there is no information at the first time, the extraction / averaging circuit 274 averages the noise power estimation values of all the subcarriers estimated by the noise power estimation circuit 273 in the frequency direction, and the noise power σ common to all the subcarriers. Estimate as _n ² . However, when the interference probability α (c) is obtained from the averaging circuit 272, the extraction / average circuit 274 determines that there is no subchannel in which the interference probability α (c) is lower than a predetermined threshold, that is, there is no interference. It is determined whether there is a subchannel that has been changed. When there is a subchannel for which it is determined that there is no interference, the extraction / averaging circuit 274 extracts only the noise power estimation values of the subcarriers belonging to the subchannel, and extracts the extracted noise power estimation values thereof. The estimation value of the noise power σ _n ² common to all the subcarriers is updated by averaging in the frequency direction between the subcarriers.

The mixed log likelihood ratio calculation circuit 275 receives the noise power estimation value σ _n ² from the extraction / averaging circuit 274 and receives the interference probability α (c) and interference power γ (c) of each subchannel from the maximization circuit 271. Enter the information. As shown in the following equation (9), the mixed log-likelihood ratio calculation circuit 275 determines the interference probability β (l) and interference power σ _if ² (l) of each subcarrier l to the subchannel c to which the subcarrier belongs. Interference probability α (c) and interference power γ (c).

The mixed log-likelihood ratio calculation circuit 275 uses the following equation (10) based on the equalized reception signal point output from the equalization circuit 263 to calculate the PSK of the transmission signal at time t, subcarrier number l or A mixed log likelihood ratio LLR ^Weigh c (t, l, m) for the m-th bit c (t, l, m) in QAM is calculated.

X ₁ (m) and X ₀ (m) are sets of signal points in which the m-th bit in modulation mapping such as PSK and QAM is 1 and 0, respectively. Mixed log-likelihood ratio calculation circuit 275 outputs mixed log-likelihood ratio ^LLR Weigh c calculated for all subcarriers (t, l, m) mixed log-likelihood ratio ^{LLR Weigh} consisting in decoding circuit 276.

The decoding circuit 276 performs error correction decoding processing on the input mixed log-likelihood ratio LLR ^Weight . The decoding circuit 276 outputs the received bit sequence obtained by the error correction decoding process to the error correction coding circuit 277. After the output of the received bit sequence from the decoding circuit 276, the outer loop is repeated.

  The error correction encoding circuit 277 outputs a transmission bit sequence obtained by performing error correction encoding similar to that of the transmitter 100 to the reception bit sequence input from the decoding circuit 276 to the switching circuit 265. The switching circuit 265 outputs the transmission bit sequence output from the error correction encoding circuit 277 to the mapping circuit 266.

Next, the operation of the receiver 200 will be described.
FIG. 4 is a flowchart showing the receiving operation of the receiver 200.
When antenna 210 of receiver 200 receives a radio signal, frequency converter 220 converts the frequency of the radio signal received by antenna 210 to obtain a received signal. The cyclic prefix removal circuit 230 removes the cyclic prefix from the frequency-converted received signal (step S105). The FFT circuit 240 performs FFT on the received signal from which the cyclic prefix has been removed to convert it to a frequency domain signal, and obtains a detection signal for each subcarrier (step S110). The P / S conversion circuit 250 rearranges the detection signals for each subcarrier in the time direction, converts the parallel signals into serial signals, and outputs them to the decoding processing unit 260.

  The reception signal input from the P / S conversion circuit 250 by the decoding processing unit 260 is branched and output to the reception signal buffer 261, the transmission path estimation circuit 262, the equalization circuit 263, and the noise power estimation circuit 273. The reception signal buffer 261 stores a reception signal.

  The transmission path estimation circuit 262 extracts a known signal from the received signal output from the P / S conversion circuit 250, estimates a channel response for each subcarrier, and obtains a transmission path estimation value from the estimated channel response for each subcarrier. (Step S115). The equalization circuit 263 equalizes the received signal for each subcarrier based on the transmission path estimation value obtained in step S115 (step S120).

On the other hand, the noise power estimation circuit 273 estimates the noise power included in the reception signal output from the P / S conversion circuit 250 for each subcarrier (step S125). The extraction / averaging circuit 274 determines that the interference parameter has not been estimated yet (step S130: NO), averages the noise power estimation values of all subcarriers estimated in step S125 in the frequency direction, and generates noise power σ. _n ² is set (step S135).

  The hard decision circuit 264 performs a hard decision on the transmission signal point candidate closest to the reception signal point on the IQ plane equalized in step S120, and estimates the transmitted bit sequence for each subcarrier (step S140). ). Hard decision circuit 264 outputs the estimated bit sequence for each subcarrier to switching circuit 265. The switching circuit 265 outputs the bit sequence estimated by the hard decision circuit 264 in step S140 to the mapping circuit 266.

  The mapping circuit 266 modulates the input bit sequence by the same modulation scheme as the transmitter 100 such as PSK / QAM, and generates a transmission replica signal (step S145). The transmission path weight circuit 267 multiplies the transmission replica signal generated in step S145 by the transmission path weight using the transmission path estimation value obtained in step S115, and generates a reception replica signal for each subcarrier (step S115). S150). The unnecessary signal power calculation circuit 268 reads the received signal from the received signal buffer 261 and subtracts the received replica signal generated in step S150 from the read received signal to obtain unnecessary signal power for each symbol. The unnecessary signal power calculation circuit 268 calculates the square value of the unnecessary signal power for each symbol (step S155).

  The subchannelization circuit 269 determines a subchannel to which each subcarrier belongs by comparing the square value of the unnecessary signal power calculated in step S155 with a threshold value (step S160). After determining the subchannel, the decoding processing unit 260 performs inner loop processing for each subchannel (step S165).

  FIG. 5 is a diagram illustrating an example of subchannelization in step S160. In the figure, the subchannelization circuit 269 uses two threshold values of threshold value E1> threshold value E2 for subchannelization. The subchannelization circuit 269 sets each subcarrier to subchannel c1 if the square value of unnecessary signal power is equal to or greater than threshold E1, subchannel c2 if threshold value E2 is less than threshold E1, and subchannel if less than threshold E2. Classify into c3. Further, the subcarrier number may be divided into a plurality of subchannels in advance instead of the determination based on the comparison between the unnecessary signal power and the threshold value. In that case, it is desirable to design the bandwidth of the subchannel so that the bandwidth is equal to the narrowest interference wave among the possible interference waves.

FIG. 6 is a flowchart showing the inner loop processing for each subchannel of the decoding processing unit 260, and shows the detailed processing of step S165 of FIG.
The log likelihood expectation value calculation circuit 270 and the maximization circuit 271 select one subchannel (step S205), and perform the following steps S210 to S225 for the selected subchannel.

  The maximization circuit 271 determines whether or not the interference parameter has converged (step S210). In the case of the first loop for the selected subchannel, the maximization circuit 271 determines that the interference parameter has not converged (step S210: NO), and performs the process of step S215.

The log likelihood expectation value calculation circuit 270 calculates the log likelihood expectation value (Q function) for each of the subcarriers belonging to the selected subchannel by using the square value equations (6) and (7) of the interference signal (step) S215). The observed data w _l, using a square value of the unwanted signal power calculated unnecessary signal power calculating circuit 268 at step S155. In the first loop, κ = 0 is substituted as an initial value, and a predetermined initial value θ ⁽⁰⁾ (l) is used.

The maximizing circuit 271 uses, for each subcarrier belonging to the selected subchannel, the interference probability β constituting the interference parameter θ ^{(κ + 1)} (l) that maximizes the Q function calculated in step S 215 according to the equation (8). ^{(Κ + 1)} (l) and interference power σ _if ^{2 (κ + 1)} (l) are calculated (step S220).

The maximization circuit 271 determines whether or not the difference between the interference parameter θ ^{(κ + 1)} (l) of the subcarrier belonging to the selected subchannel and the interference parameter θ ^(κ) (l) is within a predetermined range (step S210). ). Even if the difference between the interference probability β ^{(κ + 1)} (l) and the interference probability β ^(κ) (l) is used as the difference between the interference parameter θ ^{(κ + 1)} (l) and the interference parameter θ ^(κ) (l). Good. If the difference is not within the predetermined range, the maximization circuit 271 determines that the interference parameter has not converged (step S210: NO), adds 1 to the value of κ, and repeats the processing from step S215 to logarithmic likelihood. The expected value calculation circuit 270 is instructed. The maximization circuit 271 determines that the interference parameter has converged when the difference between the interference parameter θ ^{(κ + 1)} (l) and the interference parameter θ ^(κ) (l) is within a predetermined range (step S210: YES), the process of step S225 is performed.

When it is determined that the interference parameter has converged, the maximization circuit 271 determines the interference probability β ^{(κ + 1)} (l) and interference power σ _if ^{2 (κ + 1)} (l) of each subcarrier belonging to the selected subchannel, Output as interference parameter θ = (interference probability β (l), interference power σ _if ² (l)). The averaging circuit 272 calculates the average values of the interference probability β (l) and the interference power σ _if ² (l) of the subcarriers belonging to the selected subchannel, and the subchannel interference probability α (c) and The interference power γ (c) is calculated (step S225). The averaging circuit 272 outputs the interference probability α (c) of the selected subchannel to the extraction / averaging circuit 274, and mixes the interference probability α (c) of the selected subchannel and the interference power γ (c). The log likelihood ratio calculation circuit 275 is notified.

The log likelihood expected value calculation circuit 270 determines whether or not all subchannels have been selected (step S230). The log likelihood expectation value calculation circuit 270 repeats the processing from step S205 when there is an unselected subchannel (step S230: NO). When the log likelihood expectation value calculation circuit 270 selects all the subchannels (step S230: YES), the inner loop process ends.
The decryption processing unit 260 repeats the processing from step S130 of FIG.

In FIG. 4, the extraction / averaging circuit 274 determines that the interference parameter θ has been estimated (step S130: YES), and performs the process of step S170. That is, the extraction / averaging circuit 274 extracts only the noise power estimation value of the subcarrier to which the subchannel to which the interference probability α (c) is lower than the predetermined threshold belongs, and averages the extracted subcarriers in the frequency direction. And the estimated value of the noise power σ _n ² is updated (step S170). For example, in the example of the subchannel shown in FIG. 5, it is desirable that only the subchannel c3 be extracted. By this processing, when there is a difference between the noise power density and the interference power density, it is possible to prevent the noise power estimation value from being biased by the interference power and to improve the estimation accuracy.

The mixed log likelihood ratio calculation circuit 275 receives the noise power estimation value σ _n ² updated in step S170, and receives the interference probability α (c) and interference power γ (c) of each subchannel from the averaging circuit 272. Enter information. The mixed log likelihood ratio calculation circuit 275 uses the equation (9) to calculate the interference probability β (l) and interference power σ _if ² (l) of the subcarrier l and the interference probability α ( c) and interference power γ (c). The mixed log-likelihood ratio calculation circuit 275 calculates the mixed log-likelihood ratio LLR ^{Weight according} to equation (10) based on the equalized reception signal point output from the equalization circuit 263 (step S175).

The decoding circuit 276 performs error correction decoding processing on the mixed log likelihood ratio LLR ^Weight calculated in step S175 to obtain a hard decision bit sequence (step S180). The decoding circuit 276 determines whether or not the error correction decoding process has been performed a predetermined number of times (step S185). When determining that the error correction decoding process has not been performed a predetermined number of times (step S185: NO), the decoding circuit 276 outputs the hard decision bit sequence to the error correction encoding circuit 277. The error correction coding circuit 277 performs error correction coding similar to that of the transmitter 100 on the hard decision bit sequence obtained in step S180 to obtain a transmission bit sequence (step S190). The switching circuit 265 outputs the transmission bit sequence obtained in step S190 to the mapping circuit 266. The decryption processing unit 260 repeats the processing from step S145.

That is, the mapping circuit 266 modulates the bit sequence obtained by the error correction coding circuit 277 in step S190 using the same modulation scheme as that of the transmitter 100, and generates a new transmission replica signal (step S145). The transmission path weight circuit 267 generates a new reception replica signal by multiplying the new transmission replica signal by the transmission path weight using the transmission path estimation value (step S150). When the unnecessary signal power is updated by subtracting a new received replica signal from the received signal, the unnecessary signal power calculating circuit 268 calculates the square value of the updated unnecessary signal power for each symbol (step S155). The subchannelization circuit 269 determines a subchannel to which each subcarrier belongs by comparing the square value of the unnecessary signal power after the update and the threshold (step S160). The decoding processing unit 260 updates the observation data _wl with the updated unnecessary signal power, and performs the inner loop processing for each subchannel shown in FIG. 6 (step S165). Thereby, the decoding processing unit 260 updates the interference probability β (l) and the interference power σ _if ² (l) of each subcarrier. The averaging circuit 272 calculates the average values of the interference probability β (l) and the interference power σ _if ² (l) of the subcarriers belonging to each subchannel, and the interference probability α (c) and the interference power of the subchannel. Update γ (c).

The extraction / averaging circuit 274 determines that the interference parameter θ has been estimated (step S130: YES), and the noise power of the subcarrier to which the subchannel to which the updated interference probability α (c) falls below a predetermined threshold value belongs. The estimated values are averaged, and the estimated value of the noise power σ _n ² is updated (step S170). The mixed log-likelihood ratio calculation circuit 275 calculates the interference probability β (l) and interference power σ _if ² (l) of the subcarrier l by the equation (9), and the updated interference of the subchannel c to which the subcarrier belongs. Let probability α (c) and interference power γ (c). The mixed log-likelihood ratio calculation circuit 275 calculates a new mixed log-likelihood ratio LLR ^{Weight according} to equation (10) based on the equalized reception signal point output from the equalization circuit 263 (step S175). ). The decoding circuit 276 performs error correction decoding processing on the new mixed log likelihood ratio LLR ^Weight to obtain a new hard decision bit sequence (step S180). The decoding circuit 276 determines whether or not the error correction decoding process has been performed a predetermined number of times (step S185). When it is determined that the error correction decoding process has not been performed a predetermined number of times (step S185: NO), the decoding circuit 276 outputs a new hard decision bit sequence to the error correction encoding circuit 277. The error correction coding circuit 277 performs error correction coding similar to the transmitter 100 on the new hard decision bit sequence to obtain a transmission bit sequence (step S190). The switching circuit 265 outputs the new transmission bit sequence obtained in step S190 to the mapping circuit 266. The decryption processing unit 260 repeats the processing from step S145.
If it is determined that the error correction decoding process has been performed a predetermined number of times (step S185: YES), the decoding circuit 276 outputs the hard decision bit sequence obtained in step S180 as a received bit sequence.

  As described above, after the second outer loop process, the error correction encoding circuit 277 is encoded in the same manner as the transmitter 100 based on the received bit sequence estimated by the error correction decoding process in step S180. Generate a signal. The switching circuit 265 selects the output of the hard decision circuit 264 in the first Outer loop process, but selects the output from the error correction coding circuit 277 in the second and subsequent times. The operation after the switching circuit 265 outputs the transmission bit sequence is the same as described above, but the accuracy of the received replica signal is increased by the Outer loop process, so the accuracy of the unnecessary signal power calculated by the unnecessary signal power calculation circuit 268 is improved. To do. The subchannelization circuit 269 is also expected to improve the accuracy of observation data, which is an appropriate value of subchanneling or the input value of the EM algorithm in the inner loop processing, and to improve the estimation accuracy of interference parameters. In particular, when the number of subchannels is large, the number of subcarriers per subchannel is small and the parameter of the observation data of the EM algorithm is reduced. By improving the accuracy of the observation data by the outer loop, it is possible to expect a gain of the iterative processing. Then, after performing Outer loop processing a predetermined number of times, the output of the decoding circuit 276 is obtained as a received bit sequence.

FIG. 7 is a diagram illustrating another configuration example of the receiver according to the present embodiment. In this figure, the same parts as those of the receiver 200 shown in FIG. The receiver 200a is different from the receiver 200 shown in FIG. 3 in that a decoding processing unit 260a is provided instead of the decoding processing unit 260. The difference between the decoding processing unit 260 a and the decoding processing unit 260 is that an averaging circuit 272 a and a mixed log likelihood ratio calculation circuit 275 a are provided instead of the averaging circuit 272 and the mixed log likelihood ratio calculation circuit 275. The averaging circuit 272a receives the interference parameter θ (l) = (interference probability β (l), interference power σ _if ² (l)) from the maximization circuit 271. For each subchannel, the averaging circuit 272 calculates the average value of the interference probabilities β (l) of the subcarriers belonging to that channel, and calculates the interference probability α (c). The averaging circuit 272 outputs the interference probability α (c) of each subchannel to the extraction / averaging circuit 274, and the interference probability α (c) of each subchannel and the interference power σ _if ² (l) of each subcarrier. Is sent to the mixed log likelihood ratio calculation circuit 275a. The mixed log likelihood ratio calculation circuit 275a uses the value of the interference parameter θ (l) of each subcarrier as it is, and calculates the mixed log likelihood ratio LLR ^Weight c (t, l, m) according to the above-described equation (10). calculate.

FIG. 8 is a functional block diagram showing a configuration of a receiver 900 to which the conventional technique shown in Non-Patent Document 3 is applied. The receiver 900 can be configured as shown in the figure using a circuit similar to the receiver 200.
The decoding processing unit 960 of the receiver 900 inputs the received signal from the P / S conversion circuit 250 by the same processing as the receiver 200 and inputs it to the received signal buffer 261, the transmission path estimation circuit 262, and the equalization circuit 263. Branch and output. The reception signal buffer 261 stores a reception signal. The transmission path estimation circuit 262 obtains a transmission path estimation value from the received signal. The equalization circuit 263 equalizes the received signal for each subcarrier based on the obtained transmission path estimation value. The hard decision circuit 264 performs a hard decision on the transmission signal point candidate closest to the equalized reception signal point on the IQ plane, and estimates the transmitted bit sequence for each subcarrier. The mapping circuit 266 modulates the estimated bit sequence and generates a transmission replica signal. The transmission path weight circuit 267 multiplies the transmission replica signal by the transmission path weight using the transmission path estimation value, and generates a reception replica signal for each subcarrier. The unnecessary signal power calculation circuit 268 reads the reception signal from the reception signal buffer 261, subtracts the reception replica signal from the read reception signal, and calculates the square value of the unnecessary signal power for each symbol. Then, the log likelihood expectation value calculation circuit 970 derives a conditional expectation value (Q function) of the log likelihood for the observation data (square value of unnecessary signal power), and the maximization circuit 971 maximizes the Q function. The process of deriving the interference parameter is performed until the interference parameter converges. The mixed log-likelihood ratio calculation circuit 975 calculates a mixed log-likelihood ratio LLR ^{Weight according} to the equation (B) based on the interference parameter and the received signal point after equalization. The hard decision circuit 976 outputs a bit sequence obtained as a result of performing a hard decision on the mixed log likelihood ratio LLR ^Weight .
Thus, the receiver 900 has no subchannelization processing and can only estimate the same interference parameters (interference probability α, interference power σ _if ² ) for all subcarriers. Also, there is no outer loop, and there is no noise power estimation mechanism.

  According to the present embodiment, the receivers 200 and 200a generate a reception replica signal that is a reception signal of a desired wave replica from the reception signal, and subtract the reception replica signal of the desired wave from the reception signal to reduce the power of the unnecessary signal. Is calculated for each subcarrier. The receivers 200 and 200a classify subcarriers of unnecessary signals into subchannels. A subchannel is composed of one or more subcarriers, and is particularly effective in classification by power. The receivers 200 and 200a perform inner loop processing for operating the EM algorithm using the power of unnecessary signals as observation data for each subchannel. In the inner loop process, the receivers 200 and 200a repeatedly calculate the interference probability and the interference power at which the log likelihood expectation value is maximum. In the prior art, subchanneling is not performed.

  In addition, the receivers 200 and 200a described above are provided with Outer loop processing including error correction decoding, in addition to the Inner loop processing based on the EM algorithm. The receivers 200 and 200a use the received replica signal when calculating unnecessary signal power. The receivers 200 and 200a perform a hard decision on the transmission candidate point closest to the reception signal point after the equalization of the reception signal as the reception replica signal in the first Outer loop process, and obtained by the hard decision A bit sequence replica signal multiplied by a transmission path estimation value is used. In the second and subsequent Outer loop processing, the receivers 200 and 200a use the replica signal obtained by re-encoding the decoding result of the previous Outer loop processing multiplied by the transmission path estimation value as the received replica signal. To do. In the prior art, Outer loop processing is not performed.

Then, receivers 200 and 200a estimate the noise power for each subcarrier by calculating the variance value of preamble symbols transmitted continuously. When both the outer loop processing and the inner loop processing are the first time, the receivers 200 and 200a use a value obtained by averaging the estimated noise power values for all subcarriers as the noise power σ _n ² for all subcarriers. In other cases, an estimated value of noise power is extracted from subcarriers belonging to subchannels whose interference probabilities are below a specific threshold, and a value obtained by averaging these estimated values is used as the noise power of all subcarriers. Conventionally, it has been assumed that the noise power is known (information is obtained by another means).

According to the present embodiment, by making the number of interference waves having the same power in each sub-channel equal to or less than 1, it is possible to cope with interference waves via a plurality of interference waves and transmission lines having frequency characteristics. It becomes. Even when a plurality of interference waves arrive on the same subcarrier, the separation and detection cannot be strictly performed. However, in many cases, a specific interference wave becomes dominant, so that the interference detection accuracy is not greatly affected.
In addition, since the observation data decreases due to subchannelization, it is assumed that the EM algorithm iteration alone will reduce the estimation accuracy of interference parameters. However, error correction processing is performed in an outer loop process different from the EM algorithm, and the observation data is By sequentially increasing the accuracy, the interference parameter estimation accuracy can be improved.
In addition, by performing noise power estimation only with subcarriers that are estimated to have no interference wave, it is possible to prevent the noise power from being biased by the interference wave power, and at each repetition timing of the inner loop and outer loop. Since the noise power is sequentially updated, it can be expected that the log likelihood expectation value in the inner loop and the log likelihood ratio estimation accuracy in the outer loop will be improved.

(Second Embodiment)
FIG. 9 is a functional block diagram showing a configuration of the receiver 200b according to the second embodiment, and only functional blocks related to the present embodiment are extracted and shown. In this figure, the same parts as those of the receiver 200 of the first embodiment shown in FIG. The difference between the receiver 200b and the receiver 200 of the first embodiment is that a decoding processing unit 260b is provided instead of the decoding processing unit 260. The decoding processing unit 260b differs from the decoding processing unit 260 in that a switching circuit 265b is provided instead of the switching circuit 265, a soft decision replica generation circuit 280 is provided instead of the mapping circuit 266, and a mixed log likelihood ratio calculation circuit. The point is that a mixed log likelihood ratio calculation circuit 275b is provided instead of H.275, and a decoding circuit 276b is provided instead of the decoding circuit 276.

The mixed log likelihood ratio calculation circuit 275b does not input the interference probability α (c) and the interference power γ (c) for each subchannel from the averaging circuit 272 for the first time in the outer loop. Using the interference probability β (l) and the interference power σ _if ² (l), the mixed log-likelihood ratio is calculated by the equation (10). The mixed log likelihood ratio calculation circuit 275b operates in the same manner as the mixed log likelihood ratio calculation circuit 275 of the first embodiment in the second and subsequent outer loops.

The decoding circuit 276b performs a soft decision on the mixed log likelihood ratio calculated by the mixed log likelihood ratio calculation circuit 275b. The switching circuit 265b selects the mixed log likelihood ratio output from the mixed log likelihood ratio calculation circuit 275b for the first time in the outer loop, and selects the output of the result obtained by the decoding circuit 276b performing a soft decision after the second time. To do. Soft decision replica generation circuit 280 generates a transmission replica signal from the input likelihood by any means. For example, in the case of a QPSK signal, it is possible to generate a soft decision replica signal by the following equation (11) using 2-bit likelihoods λ ₀ and λ ₁ to be transmitted.

(Experimental result)
FIG. 10 is a diagram illustrating an experimental result of simulating the detection success probability of the receiver 200 of the first embodiment and the receiver 900 to which the conventional technique is applied. The evaluation was performed based on the case where there are two interference waves with respect to the desired wave, and these interference waves arrive at different interference intensities. The interference intensity of the first interference wave is D / U = 3 dB (decibel), and the interference intensity of the second interference wave is D / U = −3 dB. Moreover, the superposition rate of any interference wave with the desired wave is 6/64, and there is no superposition of these interference waves. The desired wave is turbo coded QPSK-OFDM transmission.

In the figure, the relationship between the detection success probability of the receiver 200 and the receiver 900 and E _b / N ₀ is represented by a graph. E _b / N ₀ is the ratio of signal power per bit to noise density. Here, the detection success probability refers to a probability when the position of the desired wave subcarrier on which the interference wave is superimposed can be completely estimated. In the receiver 900, since the interference power is constant in all bands, the detection success probability is less than 50%. On the other hand, in the receiver 200, even when a plurality of interference waves having different intensities coexist, they are clustered based on the residual power and individually detected by the EM algorithm and iterative decoding, so that E _b / N ₀ becomes higher. The detection success probability close to 1.0 is shown.

According to the embodiment described above, the receiver converts the unnecessary signal power into subchannels, and the inner loop processing that operates the EM algorithm for each subchannel, the interference probability and the interference power that maximize the log likelihood expectation value. Is repeatedly calculated. Therefore, the receiver can calculate the interference parameter regardless of the number of interference waves with respect to the desired wave and the bandwidth of the interference wave.
Further, the receiver decodes the received signal using the interference probability and the interference power obtained by the inner loop process by the outer loop process, and retransmits the decoded result to the transmission replica signal obtained by re-encoding the transmission path estimation value. Inner loop processing is performed again using the received replica signal multiplied by. Further, the receiver uses the average value of the noise power of the subcarriers included in the subchannel whose interference probability is lower than the threshold as the noise power of all the subcarriers, the received signal decoding process, the received replica signal generating process, and Used in Inner loop processing. As a result, the receiver can suppress a decrease in gain due to repetition of the EM algorithm, and can prevent the noise power from being biased by the interference wave power. In addition, it is possible to improve the accuracy of calculation of the interference probability and interference power in the inner loop processing, and to improve the estimation accuracy of the mixed log likelihood ratio in the outer loop processing.

  Some functions of the transmitter 100 and the receivers 200, 200a, and 200b in the above-described embodiments may be realized by a computer. In that case, a program for realizing a part of the functions of the transmitter 100 and the receivers 200, 200a, and 200b is recorded on a computer-readable recording medium, and the program recorded on the recording medium is stored in the computer system. You may implement | achieve by making it read in and executing. Here, the “computer system” includes an OS and hardware such as peripheral devices. The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM and a CD-ROM, and a hard disk incorporated in a computer system. Furthermore, the “computer-readable recording medium” dynamically holds a program for a short time like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line. In this case, a volatile memory inside a computer system serving as a server or a client in that case may be included and a program held for a certain period of time. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

  The embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and includes designs and the like that do not depart from the gist of the present invention.

  The present invention can be used for an apparatus that receives a signal transmitted by multicarrier.

1 wireless communication system 100 transmitter (transmitter)
110 Error correction coding circuit 120 Mapping circuit 130 S / P conversion circuit 140 IFFT circuit 150 Cyclic prefix addition circuit 160 Frequency converter 170 Antenna 200, 200a, 200b Receiver (receiving device)
210 Antenna 220 Frequency converter 230 Cyclic prefix removal circuit 240 FFT circuit 250 P / S conversion circuits 260, 260a, 260b Decoding processing unit 261 Reception signal buffer 262 Transmission path estimation circuit 263 Equalization circuit (replica signal generation unit)
H.264 hard decision circuit (replica signal generator)
265, 265a, 265b switching circuit 266 mapping circuit (replica signal generation unit)
267 Transmission path weight circuit (replica signal generator)
268 Unnecessary signal power calculation circuit (unnecessary signal power calculation unit)
269 Subchannelization circuit (subchannelization section)
270 Log likelihood expectation value calculation circuit (estimator)
271 Maximization circuit (estimator)
272, 272a Averaging circuit (estimator)
273 Noise power estimation circuit 274 Extraction / averaging circuits 275, 275a, 275b Mixed log likelihood ratio calculation circuit (mixed log likelihood ratio calculation unit)
276, 276b decoding circuit 277 error correction coding circuit 280 soft decision replica generation circuit

Claims (8)

A receiving apparatus that receives a signal in which a desired wave transmitted using a multicarrier superimposed transmission scheme and one or more interference waves that interfere with the desired wave are superimposed,
A replica signal generation unit that generates a replica signal of the received signal of the desired wave using a transmission signal estimated from the received signal;
An unnecessary signal power calculation unit that subtracts the replica signal from the received signal to calculate the power of unnecessary waves for each subcarrier;
A subchannelizer for determining a subchannel to which the subcarrier belongs;
The interference probability, which is the probability that an interference wave exists in the desired wave, and the interference power of the interference wave are set for each subcarrier so that the expected value of the likelihood for the power of the unnecessary wave of each subchannel is maximized. An estimation unit for estimation;
A receiving apparatus comprising: The estimation unit averages the interference probability and the interference power estimated for the subcarriers belonging to the subchannel to estimate an interference probability and interference power for each subchannel.
The receiving apparatus according to claim 1. After calculating the noise power common to all subcarriers based on the known signal included in the received signal, and obtaining the interference probability for each subchannel, the average interference probability is below a predetermined threshold A noise power estimation unit that updates noise power common to all subcarriers based on noise power of the subcarriers belonging to a subchannel;
The receiving device according to claim 2. The subchannelization unit determines a subchannel to which the subcarrier belongs according to the power level of the subcarrier in the unnecessary wave.
The receiving apparatus according to any one of claims 1 to 3, wherein the receiving apparatus is characterized in that Further comprising a mixed log likelihood ratio calculation unit that calculates a mixed log likelihood ratio from the received signal using the interference probability and the interference power estimated by the estimation unit,
The receiving device according to any one of claims 1 to 4, wherein the receiving device is configured as described above. A process in which the replica signal generation unit generates a new replica signal from a result of performing error correction decoding by updating the mixed log likelihood ratio with the interference probability and the interference power estimated by the estimation unit; ,
The unnecessary signal power calculation unit updates the unnecessary wave power by subtracting the new replica signal from the received signal;
The subchannelization unit determines a subchannel to which the subcarrier in the unnecessary wave belongs;
The estimation unit repeatedly performs the process of estimating the interference probability and the interference power based on the updated power of the unnecessary wave,
The receiving apparatus according to claim 5. The replica signal is a signal obtained by hard-decision to a candidate of a transmission signal point closest to the reception signal point of the reception signal, and the new replica signal is updated to the updated mixed log likelihood ratio. A signal obtained by performing error correction coding on the result of error correction decoding, or the replica signal is a mixed logarithmic likelihood calculated from the received signal using an interference probability and interference power of an initial value for each subcarrier. It is a soft decision replica signal based on a degree ratio, and the new replica signal is a signal generated from a result obtained by performing a soft decision on the updated mixed log likelihood ratio.
The receiving apparatus according to claim 6. A reception method executed by a receiving apparatus that receives a signal in which a desired wave transmitted using a multicarrier superimposed transmission method and one or more interference waves that interfere with the desired wave are superimposed,
A replica signal generating step of generating a replica signal of the received signal of the desired wave using a transmission signal estimated from the received signal;
Unnecessary signal power calculation step of subtracting the replica signal from the received signal to calculate the power of unnecessary waves for each subcarrier;
A subchannelization step for determining a subchannel to which the subcarrier belongs;
The interference probability, which is the probability that an interference wave exists in the desired wave, and the interference power of the interference wave are set for each subcarrier so that the expected value of the likelihood for the power of the unnecessary wave of each subchannel is maximized. An estimation step to estimate;
An interference estimation method characterized by comprising:

Images