KR100204593B1 - The decoding apparatus at fading channel in mobile communication system - Google Patents

Abstract

In order to cope with fading channel environment with intersymbol interference in mobile communication, error propagation of equalizer occurs in receiver using decision feedback equalizer and convolutional decoder, which greatly reduces the performance of system. Therefore, there is a need to develop a decoding apparatus that is robust against such error propagation. Conventionally, in order to cope with error propagation of a decision feedback equalizer, a form in which a decision feedback equalizer, a deinterleaver, and a decoder are combined into one process is considered, but the complexity of the system is greatly increased and the performance is not greatly improved.

Therefore, in the present invention, when the mobile communication channel undergoes deep fading and intersymbol interference, the error of the hard decision value of the equalizer output of the decision feedback equalizer occurs once, the probability of error propagation of the equalizer is very high. A decoder device robust to error propagation of an equalizer is presented.

By estimating the equalizer output reliability coefficients for obtaining the confidence level of the output value of the equalizer and using them as decoding metrics in Viterbi decoding, the system performance is greatly improved.

Description

Decoding Device in Mobile Fading Channel

In order to cope with fading channel environment with intersymbol interference in mobile communication, mobile communication fading robust to error propagation of equalizer in case of using decision feedback equalizer (DFE: hereinafter referred to as DFE) and decoder A decoding apparatus in a channel.

The present invention belongs to the radio wave signal processing technology in the wireless access technology field of mobile communication, and conventionally combines a decision feedback equalizer, a deinterleaver, and a decoder in one process to cope with error propagation of the decision feedback equalizer. I think, but it has the disadvantage that the complexity of the system is greatly increased and the performance is not greatly improved.

Accordingly, an object of the present invention is to provide a decoding apparatus in a mobile communication fading channel which is very simple and greatly improved in order to improve the complexity of the system in the conventional manner and the performance is not greatly improved.

In order to achieve the above object, the present invention provides equalizer output reliability estimation means for determining the degree of reliability of the output value of the equalizer between the decision feedback equalizer and the deinterleaver of the receiver in order to be robust to error propagation of the equalizer. It is characterized by.

1 is a block diagram of a decoding apparatus in a conventional mobile communication fading channel.

2 is a block diagram of a decoding apparatus in a mobile communication fading channel according to the present invention.

3 is a transmission symbol sequence frame structure diagram.

* Explanation of symbols for main parts of the drawings

101: Convolution coder 102: Interleaver

103: insert training symbol 104: transmission filter

105: transmit antenna 106: receive antenna

107: crystal feedback equalizer (DFE) 108: DFE front filter

109: DFE rear filter 110: DFE hard decision

111: deinterleaver 112: Viterbi decoder

201: Convolution Coder 202: Interleaver

203: Insert training symbol 204: Transmission filter

205 transmit antenna 206 receive antenna

207: crystal feedback equalizer (DFE) 208: DFE front filter

209 DFE rear filter 210 DFE hard decision

211: equalizer output reliability calculation unit 112: deinterleaver

213: Viterbi Decoder

The use of an equalizer can be considered to cope with Intersymbol Interference (ISI) caused by multiple waves in mobile communication. For frequency selective fading channels with deep nulls, nonlinear equalizers are efficient in terms of system performance. Decision Feedback Equalizer (DFE), a representative non-linear equalizer, is used, and a convolutional coder is used as a channel encoder to improve the performance of the system in a mobile communication environment where fading exists. Think of a transceiver.

Figure 1 shows a block diagram of a conventional system having a decision feedback equalizer and a convolutional encoder in a frequency selective fading channel. The transmitter convolutionally encodes the information to be sent by the convolutional encoder 101, uses the interleaver 102 to disperse the continuous errors occurring in the channel, and uses a training symbol insertion slot to equalize the ISI generated in the channel. In step 103, the training symbol string is inserted at a predetermined period (per transmission frame) and then output through the transmission antenna via the transmission filter 104. The receiver obtains an initial filter coefficient of the DFE 107 using the training symbol sequence transmitted from the DFE 107, and adjusts the filter coefficient of the DFE 107 using the filter coefficient value and the hard decision 110 output of the equalizer. Continue until the end of the transmission frame. The output of the equalizer is sent from the deinterleaver 111 to the Viterbi decoder 112 after the reverse process of the interleaver 102. The Viterbi decoder 112 corrects an error generated in the transmission process to finalize the transmission information bit. Will be decided. The reason for using the interleaver and deinterleaver above is as follows. Convolutional codes have the ability to correct errors effectively when they appear sporadically, but when the errors occur continuously, performance degrades significantly. The mobile communication channel has a deep fading phenomenon, so that errors in the transmission signal are continuously generated. Therefore, performance improvement is difficult even with convolution codes. Therefore, the interleaver 102 and the deinterleaver 111 have been used to improve the performance by passing the continuous error to the Viterbi decoder 112 as a sporadic error. In Viterbi decoder 112, if the metric value of the decoder is calculated using the soft decision result rather than the hard decision result of the equalizer output, the improved system performance can be seen. It is well known.

The crystal feedback equalizer having a TDL (tapped delay line) structure is composed of a front filter 108, a rear filter 109, and a hard decision portion 110, as in the DFE 107 of FIG. 1. The front filter input is a received signal and the rear feedback filter input is a previously determined symbol. The function of the rear filter removes the interference terms caused by the symbols previously determined from the prediction of the current symbol. This equalizer is equalizer And hard decision Equalizer under the assumption And hard decision Since the DFE 107 coefficient is adjusted so that the error e (K) is minimized, the hard decision value is wrong, resulting in a decrease in the reliability of the DFE 107 output.

The decision feedback equalizer usually generates error propagation for all subsequent symbols when the hard decision value to be fed back is incorrectly determined when the modulation symbols have different sizes, such as M-QAM, in a fading environment such as mobile communication. Significantly reduces the performance. In order to solve this problem, the present invention devised a scheme for improving the performance of the system by using the soft decision output and hard decision output information of the equalizer.

The block diagram of the system for improving the problem of error propagation is shown in FIG.

In FIG. 2, an equalizer output reliability estimate 211 is added between the DFE 107 and the deinterleaver 111 of FIG. The reliability estimation value is input to the Viterbi decoder 213 along with the DFE 207 output signal through the deinterleaver 212, and the Viterbi decoder 213 performs decoding using the reliability coefficient and the deinterleaver output value of the signal. Perform. The transmission frame format is shown in FIG.

The process is described in detail as follows. Soft decision output of equalizer of system is I (K), hard decision thrust , Wow When the error between them is called e (k) , they are given by Equation 1 below.

Here, i denotes an in-phase component, q denotes a quadrature component, and k is a time index in symbol units of an information symbol string in one transmission frame.

In the case where the minimum distance between constellation points of the signal is d, it may be considered that a decision error occurred when the magnitude of the in-phase component error or the quadrature component error is larger than d / 2. In the case of using a modulator with different signal levels like M-QAM in a fading environment, if the hard decision value of the decision feedback equalizer occurs once, the probability of nearly continuous error propagation is very high. The confidence level of the soft decision value of the equalizer can be estimated by looking at the above error signal (performed by the equalizer reliability estimation). If the value indicating this is called a reliability coefficient, this reliability coefficient is used as a metric weight in decoding (performed by a Viterbi decoder). The reliability factor of the annual determination of the equalizer is In the case of, the reliability of all the symbols from the symbol at that moment to the end of the information frame is set to 0, otherwise, it is set to 1 (performed by equalizer reliability estimation). That is, symbol strings having a high probability of error propagation are not allowed to participate in the decoding process of the decoder. This can be represented by the following [Equation 2].

Here, N is the number of symbols in one information symbol string.

The reliability coefficient and the equalizer soft decision output value are sent to the deinterleaver 212, and the Viterbi decoder 213 performs decoding using the two outputs from the deinterleaver 212. At the decoder, the path metric M is obtained by adding a modified branch metric given by Equation 3 along an arbitrary path (performed by the Viterbi decoder).

here Is the transmission symbol corresponding to the nth branch in a given path. And w '(n) are the soft decision output and the reliability coefficient of the equalizer through the deinterleaver, respectively. L is the decision block size of the decoder, that is, the decision depth.

As described above, the present invention utilizes an equalizer output reliability coefficient that calculates the reliability of the output value of the equalizer in order to be robust to error propagation of the equalizer in a mobile communication channel that suffers from deep fading and intersymbol interference. It is used as Weisht of decoding metric at the time, and it has the effect of greatly improving system performance.

Claims (1)

A decoding apparatus comprising a transmitter comprising a convolutional encoder, an interleaver, and a training symbol string insertion function in a frequency selective fading channel, and a receiver comprising a decision feedback equalizer, a deinterleaver, and a Viterbi decoder, so as to be robust against error propagation of the equalizer. And an equalizer output reliability estimating means for obtaining a degree of confidence of an output value of the equalizer between the decision feedback equalizer and the deinterleaver of the receiver.