JP2012124879A - Receiving device and receiving method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a receiving device and a receiving method enabling real-time and high-accuracy estimation of a BER (Bit Error Rate) value.SOLUTION: A transmitting device 10 has: a modulator 12 mapping data 11 on a modulation symbol to generate a data symbol; and a CP inserting part 13 inserting a cyclic prefix into the data symbol. The generated transmission signal is transmitted via a transmission antenna 14. A receiving device 20 has: a CP removing part 22 removing the CP from the received signal received via a receiving antenna 21; an S/P converter 23 performing serial-to-parallel conversion; an FFT part 24 performing time-to-frequency conversion; a communication path estimating and frequency region equalizing part 25 performing communication path estimation and frequency region equalization based on a received pilot signal; an error rate estimating part 26 estimating an error rate based on a communication path estimation value; an IFFT part 27 performing frequency-to-time conversion; a P/S converter 28 performing parallel-to-serial conversion; and a demodulator 29 performing demodulation. The receiving device 20 generates reception data 30.

Description

  The present invention relates to a data transmission system in a digital wireless communication system. In particular, for single-input single-output (SISO) systems and multi-input, multi-output systems for frequency-selective communication channels for PSK signals using the SC-FDE method. The present invention provides a receiving apparatus and a receiving method that realizes a highly accurate bit error rate estimation by realizing a system (Multiple-Input Multiple-Output, MIMO, hereinafter referred to as MIMO) with a simple transceiver configuration.

  In a mobile communication environment such as inter-vehicle communication or road-to-vehicle communication, it is necessary to transmit and receive reliable data in real time without retransmission. If the receiver can know the BER value for each packet during communication, it is possible to flexibly use UEP (Unequal Error Protection) transmission that ensures high reliability by switching the modulation / demodulation method according to the channel environment. It is.

  As a first conventional technique, there is frequency domain equalization (hereinafter referred to as FDE). Frequency equalization and signal separation are simultaneously performed using a method called Nulling in which received data is transformed into the frequency domain by Fast Fourier Transform (FFT) on the receiver side and multiplied by a weight matrix based on MMSE. After this processing, the data is demodulated by returning to the time domain by inverse fast Fourier transform (IFFT) (see Non-Patent Document 1).

  As a second conventional technique, there is a Per Subcarrier Estimation (PSE) method which is a communication path measurement method. The pilot signals that are known to each other are communicated by the transceiver, and the reception pilot value in the frequency domain is divided by the transmission pilot value to obtain the channel measurement value (Non-patent Document 2).

Utsunomiya, Takahashi, Iwanami, Okamoto, “A Comparative Study on LDPC Coded MIMO SC FDE Systems”, IEICE Society Conference B-5-56, September 2007 D. Takeda, Y. Tanabe, “Channel Estimation Scheme with Low-Complexity Discrete Cosine Transform in MIMO-OFDM,” IEICE transactions on Communications E92.B, No.12, pp.3836-3842, December 2008

  However, in the above-described conventional technology, since the frequency selective communication channel in mobile communication changes with time, it is difficult to estimate BER for each packet by estimating BER by demodulating a pilot signal. In addition, in order to perform BER estimation with sufficient accuracy, a huge number of known signals are transmitted, and transmission efficiency is deteriorated.

  In view of such circumstances, the present invention intends to provide a receiving apparatus and a receiving method that realize real-time and highly accurate estimation of a BER value.

  The present invention uses a pilot signal to estimate a channel, and uses a channel estimation / frequency domain equalization unit that estimates a channel estimation value and average noise power, and uses the channel estimation value and the average noise power. And an error rate estimator for calculating the error rate.

  Here, the error rate estimation unit estimates the error rate by approximating the communication path after frequency domain equalization to AWGN.

  The error rate estimation unit estimates an error rate for each transmission antenna in the MIMO communication system.

  Further, the present invention provides a channel estimation / frequency domain equalization process for estimating a channel using a pilot signal and estimating a channel estimation value and average noise power, and the channel estimation value and the average noise power. And an error rate estimation process for calculating an error rate using the.

  As described above, in the SC-FDE method, the present invention calculates the gain and noise power of the equivalent AWGN channel from the measured channel characteristic value and the average noise power measurement value, thereby realizing a real-time and high-accuracy BER. Can be estimated.

It is a figure which shows the system model of the system of a SISO transmission / reception apparatus. It is a figure which shows FDE and impulse response in ZF standard. It is a figure which shows FDE and an impulse response in a MMSE standard. It is a model figure which shows the outline of a multipath channel model. In a SISO system, it is the figure which showed the comparison result of the BER characteristic by computer simulation in case channel measurement is perfect at the receiving side. It is the figure which showed the comparison result of the BER characteristic by the computer simulation at the time of performing a channel measurement in a SISO system. It is the figure which showed the comparison result of the BER characteristic by the computer simulation in case a channel measurement is perfect in the receiving side in a 2x2 MIMO system. It is the figure which showed the comparison result of the BER characteristic by the computer simulation at the time of measuring a channel in a 2 * 2 MIMO system. It is the figure which showed the comparison result of the BER characteristic by the computer simulation in case a channel measurement is perfect at the receiving side in a 4x4 MIMO system. It is the figure which showed the comparison result of the BER characteristic by the computer simulation at the time of performing a channel measurement in a 4x4 MIMO system. It is a figure which shows the system model of the system of a MIMO transmission / reception apparatus.

  Embodiments of the present invention will be described below with reference to the accompanying drawings.

(First embodiment)
First, the present embodiment will be described with reference to FIG. FIG. 1 shows the case of a SISO system.
The transmission apparatus 10 includes a modulation unit 12 that maps data 11 to modulation symbols to generate data symbols, and a CP insertion unit 13 that inserts cyclic prefixes into the data symbols. 14 is transmitted.

  The receiving apparatus 20 includes a CP removing unit 22 that removes a CP from a received signal received by the receiving antenna 21, an S / P converting unit 23 that performs serial / parallel conversion, an FFT unit 24 that performs time-frequency conversion, and a communication path from the received pilot signal. A channel estimation / frequency domain equalization unit 25 that performs estimation and frequency domain equalization, an error rate estimation unit 26 that estimates an error rate from a channel estimation value, an IFFT unit 27 that performs frequency-time conversion, and a P / that performs parallel-serial conversion An S conversion unit 28 and a demodulation unit 29 that performs demodulation are provided, and receive data 30 is generated.

  The transmission apparatus 10 inserts a Cyclic Prefix (CP) into the data symbol x (k) for transmission. The transmission signal propagates through the frequency selective fading channel and is received by the receiving device 20. The receiving apparatus 20 removes CP from the received signal, and then decomposes it into N frequency components {R (i), i = 0, 1,..., N−1} by N-point FFT. R (i) is represented by the following formula (1).

  In equation (1), r (k) is the kth time sample value. H (i) represents channel characteristics in the i-th frequency sample, and X (i) and Z (i) represent the i-th frequency component of the transmission signal and noise, respectively. When each frequency component is multiplied by the FDE weight W (i) based on the MMSE standard, the sample value r ′ (k) of the time domain signal after FDE is expressed by the following equation (2).

    W (i) is the following equation (3).

  In equation (2), h '(k), which is the kth time sample value, is an impulse response that combines the channel and the equalizer, and is represented by the following equation (4).

  In the ZF-based FDE, since W (i) H (i) = 1 at the i-th frequency, this impulse response h ′ (k) is {h ′ (k) = 0, k ≠ 0}. (FIG. 2). However, in the FDE of the MMSE standard, W (i) H (i) = 1 is not necessarily set, and therefore a value also exists in a section where k ≠ 0 in h ′ (k) (FIG. 3). This affects signal determination as intersymbol interference.

  The expression (2) is modified as the following expression (5) in order to use the coefficient for the data symbol x (k) as a gain.

Since the coefficient relating to the data symbol x (k) is defined as a gain, the gain A _eq of the equivalent AWGN is expressed by the following equation (6).

By approximating the second term of equation (4) to a Gaussian distribution, the variance of the combination of the second and third terms of equation (4) is calculated, and this is calculated as the average noise power σ ² _eq of the equivalent AWGN Then, it is represented by the following formula (7).

In the above, the channel value H (i) and the average noise power value σ ² are values that can be measured using pilot signals that are known to each other by the transceiver. If the gain A _eq and noise variance σ ² _{eq of} the equivalent AWGN can be calculated in advance using the channel value H (i) and the average noise power value σ ² , the BPSK modulation or the QPSK modulation can be performed using the Q function. The received BER can be estimated from the following equation (8).

Now, it is necessary to obtain the channel value and the average noise power by the receiving device, which will be described in detail below.
In order to obtain the channel value, a known pilot signal may be used in the receiving apparatus. For example, assuming that a pilot signal is allocated to each subcarrier, the channel estimation value H ^ (i) at the i-th frequency point is expressed by the following equation (9) using the pilot signal P (i) at the i-th frequency point. It is required as follows.

However, since noise is included in R (i), a method for suppressing the influence of noise and measuring noise power will be described. ^First, determined by ^H ~ (i) an in IFFT impulse response estimate ^h ~ (k) the following equation (10).

Here, h (k) is an actual channel impulse response. Z ′ (k) is complex Gaussian noise. If the average power of P (i) is 1, the average power of z (k) and z ′ (k) is equal. Further, since h (k) has a value only up to the sample number L-1, the values after the (L-1) th of h ^to (k) are only the values of z '(k). By setting the value to 0 and then performing FFT, a channel estimation value H ^ (i) in which the influence of noise power is suppressed to L / N as shown in the following equation (11) is obtained.

Moreover, the measured value σ ^ ² of the average power of noise can be measured from the average value after the Lth of h ^to (k) as in the following equation (12).

  In order to increase the accuracy of error rate estimation, it is necessary to obtain a channel estimation error. The relational expression between the communication channel estimated value H ^ (i) and the actual communication channel H (i) is expressed as the following equation (13).

However, Z ^~ (i) is a channel estimation error. When the equalized signal is obtained in consideration of the above equation (13), the following equation (14) is obtained.

z ″ (k) is obtained by adding the noise of the channel estimation value to z ′ (k). Assuming that z ″ (k) is Gaussian distributed, the final average noise power of the equivalent AWGN channel is expressed by the following equation (15).

(Second Embodiment)
The present embodiment is an example in which the present invention is applied to SC-FDE in a MIMO (Multiple Input Multiple Output) system having a plurality of antennas on the transmission / reception side. It is assumed that the number of transmission antennas is n _T and the number of reception antennas is n _R.

FIG. 11 is a block diagram illustrating a configuration of a transmission device and a reception device in the present embodiment.
The transmission apparatus 40 includes modulation units 42-1 to 42-n _T that map data 41-1 to 41-n _T to modulation symbols to generate data symbols, and CP insertion units 43 that insert cyclic prefixes into the data symbols. -1 to 43-n _T and a plurality of generated transmission signals are transmitted from the plurality of transmission antennas 44-1 to 44-n _T.

Receiving apparatus 50, the receiving antennas 51-1 to _51-n removes CP from the reception signal received by _R CP removing section _{52-1~52-n R,} S / P converter for serial-parallel conversion 53- 1 to _{53-n R,} FFT unit 54-1 to _{54-n R,} channel estimation and frequency domain equalization unit 55 for channel estimation and frequency domain equalization from the received pilot signals to time-frequency conversion, the channel An error rate estimator 56 that estimates an error rate from an estimated value, an IFFT unit 57-1 to 57-n _R that performs frequency-time conversion, a P / S converter 58-1 to 58-n _R that performs parallel-serial conversion, and a demodulator Demodulating units 59-1 to 59-n _{R for} performing reception data 60-1 to 60-n _R.

Hereinafter, detailed description will be given using mathematical expressions.
The n _R- dimensional received signal vector R (i) at the i-th frequency point is represented by the following equation (16).

Where R _q (i) is a received signal at the qth receiving antenna, Xp (i) is a transmitted signal at the pth transmitting antenna, and H _qp (i) is a communication path between the pth transmitting antenna and the qth receiving antenna. , Z _q (i) is noise in the q-th receiving antenna. Superscript T represents a transposed matrix.

  In a MIMO system, a non-linear detection method such as MLD (Maximum Likelihood Detection), ZF (Zero Forcing) or MMSE (Minimum Mean Square Error) is used to separate and detect transmission signals from received signals. A linear detection method such as Hereinafter, the linear detection method will be described. In the case of the linear detection method, the received signal is detected by separating it by multiplying the received signal by a weight matrix W (i) as shown in the following equation (17).

  The weight matrix W (i) is expressed by the following equation (18) in the case of MMSE.

  In the case of ZF, the following equation (19) is obtained.

  Note that IN represents a unit matrix of N rows and N columns. Superscript H represents a complex conjugate transpose matrix. The time domain signal x ^ (k) of X ^ (i) is expressed by the following equation (20).

  The impulse response h ′ (k) after separation detection is expressed as in the following equation (21):

  x ^ (k) is expressed by the following equation (22).

  When attention is paid to the m-th transmitting antenna and the signal at time k, the following equation (23) is obtained.

  At this time, the equivalent AWGN gain at the transmitting antenna m and time k is expressed by the following equation (24).

  The average noise power is as shown in the following equation (25).

  In the MIMO system, an inter-stream interference component is newly added as noise in the case of the SISO system.

(Third embodiment)
In the present embodiment, the present invention is applied to an OFDM (Orthogonal Frequency Division Multiplexing) scheme.

In the case of OFDM, an error rate can be obtained for each subcarrier, and the average can be the overall error rate. The frequency domain received signal and equalization weight may be the same as in SC-FDE. Therefore, the equivalent gain A _eq (i) and the average noise power σ ² _eq (i) of the i-th subcarrier are expressed by the following equation (26).

  If the error rate of each subcarrier is obtained using the above equation and averaged over all subcarriers, the overall error rate can be estimated.

(Fourth embodiment)
In the present embodiment, the present invention is applied to a MIMO-OFDM system in which OFDM is extended to MIMO.

The received signal and the weight matrix in the i-th subcarrier are the same as in MIMO-SC-FDE. At this time, the equivalent AWGN gain A _{eq, m} (i) of the m-th transmission antenna and the i-th subcarrier is expressed by the following equation (27).

Further, the average noise power σ ² _{eq, m} (i) of the m-th transmission antenna and the i-th subcarrier is expressed by the following equation (28).

However, inter-stream interference approximates noise. The error rate of each subcarrier is obtained using such A _{eq, m} (i) and σ ² _{eq, m} (i) and averaged over all subcarriers, so that the overall error rate can be estimated.

(Fifth embodiment)
Here, a case where pilot signals are arranged in the time domain will be described. The channel value measurement method can be performed in the same manner as in the first embodiment.
When a pilot signal is arranged in the time domain, the pilot signal does not have a constant amplitude in the frequency domain and changes for each frequency point. At this time, when the channel estimation value H ^ (i) at the i-th frequency point is obtained, if the amplitude of the pilot signal at the i-th frequency point is small, noise enhancement occurs and the channel estimation accuracy deteriorates. In an extreme case where the amplitude of the pilot signal is 0, the channel estimation itself cannot be performed. For this reason, it is necessary to use a pilot signal that reduces the channel measurement error. At this time, the channel estimation error can be expressed by the following equation (29), and a pilot signal that makes equation (29) small may be used. Further, P (i) in the following equation (30) is a pilot signal at the i-th frequency point. For example, a pilot signal with a small estimation error may be selected by selecting a pilot signal that minimizes Equation (29) from among a plurality of pilot signal candidates.

(Sixth embodiment)
In the case of a MIMO system, the number of communication paths to be estimated increases in proportion to the number of transmission antennas. If an attempt is made to estimate these communication paths using independent resources (symbols), the pilot signal for the data signal increases as the number of transmission antennas increases, leading to deterioration in transmission efficiency. A technique for performing MIMO channel estimation using the same resource will be described.

  For example, CI (Carrier Interferometry) technology can be used. In the CI technique, the amplitude and phase of all subcarriers are made the same, and a linear phase offset is given to each subcarrier as shown in the following equation (31).

N _p represents a phase offset coefficient and changes the slope of the straight line. Giving a linear phase offset in the frequency domain results in a time shift according to N _p in the time domain. Even if the pilot signals of the antenna are multiplexed, the channel estimation value can be obtained with high accuracy. Also, when pilot signals are multiplexed in the time domain, providing a cyclic delay with a different delay amount at each transmitting antenna is equivalent to providing a linear phase offset in the frequency domain. For example, when [s1 s2 s3 s4] is cyclically delayed by a delay amount 1, [s2 s3 s4 s1] is obtained.

  It is also possible to perform channel estimation by performing frequency multiplexing by distributing pilot signals of each transmitting antenna. The interval of the distributed arrangement is preferably equal, and the interval of the distributed arrangement increases as the number of transmission antennas increases. A case of 8 subcarriers will be described as an example. For example, when the number of transmission antennas is 2, the first transmission antenna places pilot signals on the first, third, fifth and seventh subcarriers, and the second transmission antenna places pilot signals on the second, fourth, sixth and eighth subcarriers. Place. In order to obtain a channel estimation value of a subcarrier on which no pilot signal is arranged at each transmission antenna, for example, interpolation may be performed using the pilot signal of the arranged subcarrier.

  It is also possible to combine CI technology and frequency multiplexing technology. By multiplexing the pilot signals with a linear phase offset added to the distributed pilot signals, even if the number of transmission antennas increases, the interval of the distributed arrangement can be reduced, and the number of pilot signals arranged at each antenna increases. Communication path estimation accuracy can be improved.

  Thus, in the said embodiment, the method of estimating BER accurately was shown. This makes it possible to avoid transmission errors by discarding packets with a high error rate (or selecting whether or not to perform subsequent reception processing). Also, in the case of a MIMO system, highly reliable data transmission can be performed for each transmission antenna by discarding a packet with a high error rate for each transmission antenna. Further, it may be determined whether or not all transmission antennas are discarded together. In addition, it is desirable to select whether or not to discard the criteria, for example, depending on QoS. The criteria may be stricter for data with high importance, and the criteria may be set low for data with low importance. In an environment where the error rate is poor, the transmission rate may be lowered in order to increase reliability. This may be determined by the transmission device, or the reception device may notify the transmission device.

(Example)
Hereinafter, although the effect of the present invention is explained concretely based on an example, the present invention is not limited to these examples from the first.

  The channel between each transmitting / receiving antenna is a frequency selective quasi-static Rayleigh fading channel of equal power 1, 2, 4, 8, 16 symbol time paths independent of each other. . The delay profile of this channel is shown in FIG.

  One block of FFT includes 64 symbols (that is, N = 64). The GI (Guard Interval) length is N / 4 points. The channel estimation of the communication channel is compared between the case where the reception side is perfect and the case where measurement is performed using a pilot signal.

  For a SISO system with 1 × 1 antennas, when the channel estimation is complete on the receiving side in FIG. 5 and the channel is measured with a pilot signal in FIG. 6, the error rate (Bit Error Rate: BER) characteristics by computer simulation Indicates. In all cases, it is implemented for SISO systems and 2 × 2, 4 × 4 MIMO systems. 5 to 10, the reception BER value was estimated by measuring the channel information (CSI) and the average reception noise power. Based on the channel information and reception noise power measured by the transmission pilot signal, the gain and noise variance of the equivalent AWGN channel can be calculated, and the theoretical expression of the reception BER value can be obtained, and BER estimation with high accuracy is possible. It was confirmed that.

The present invention relates to a data transmission system in a digital wireless communication system. In particular, for the SC-FDE scheme, real-time and high-precision BER estimation is possible, and UEP (Unequal Error Protection that ensures high reliability by switching the demodulation scheme according to the reception E _b / N ₀ environment. ) There is a possibility that it can be applied flexibly to transmission. Further, the MIMO system for BER estimation can be implemented with a simple configuration, and the applicability to a MIMO digital wireless communication system that realizes an improvement in transmission speed and excellent bit error rate characteristics is great.

DESCRIPTION OF SYMBOLS 10 Transmission apparatus 11 Data 12 Modulation part 13 CP insertion part 14 Transmission antenna 20 Reception apparatus 21 Reception antenna 22 CP removal part 23 S / P conversion part 24 FFT part 25 Communication path estimation / frequency domain equalization part 26 Error rate estimation part 27 IFFT unit 28 P / S conversion unit 29 Demodulation unit 30 Received data 40 Transmitting devices 41-1 to 41-n _T data 42-1 to 42-n _T modulation units 43-1 to 43-n _T insertion units 44-1 to _{44-n T} transmit antennas 50 receiving apparatus 51-1 to _{51-n R} receiving antennas 52-1 to _52-n R CP removing section _53-1~53-n R S / P conversion section 54-1 to 54-n _R FFT unit 55 Channel estimation / frequency domain equalization unit 56 Error rate estimation unit 57-1 to 57-n _R IFFT unit 58-1 to 58-n _R P / S conversion unit 59-1 to 59-n _R demodulation Part 60-1 to 60- _R received data

Claims (4)

A channel estimation unit that estimates a channel using a pilot signal and estimates a channel estimation value and average noise power; and
An error rate estimator that calculates an error rate using the channel estimation value and the average noise power;
A receiving apparatus comprising:   The receiving apparatus according to claim 1, wherein the error rate estimation unit estimates an error rate by approximating a communication path after frequency domain equalization to AWGN.   The receiving apparatus according to claim 1, wherein the error rate estimation unit estimates an error rate for each transmission antenna in a MIMO communication scheme. Estimating the channel using the pilot signal, channel estimation and frequency domain equalization process to estimate the channel estimate and average noise power,
An error rate estimation process of calculating an error rate using the channel estimation value and the average noise power;
A receiving method comprising:

Images