13-10-2010, 02:51 PM
[attachment=5879]
This article is presented by:
Xiaoyan Xu
Mark F. Flanagan
Member, IEEE
Norbert Goertz
Senior Member, IEEE
John Thompson
Joint Channel and Network Coding for Cooperative Diversity in a
Shared-Relay Environment
Shared-Relay Environment
ABSTRACT
In this paper we propose a cooperative diversity scheme for the communication model of two sources sharing a single relay. The scheme uses algebraic code superposition relaying in the multiple access fading channel to create spatial diversity under the constraint of limited communications resources. We also describe in detail a novel computationally efficient message passing algorithm at the destination’s decoder which extracts the substantial spatial diversity contained in the code superposition and signal superposition. The decoder is based on a sliding window structure where certain a posteriori LLRs are retained to form a priori LLRs for the next decoding. We show that despite the simplicity of the proposed scheme, diversity gains are efficiently leveraged by the simple combination of channel coding at the sources and network coding at the relay.
INTRODUCTION
IT is well known that spatial diversity can effectively combat the deleterious effect of fading . In recent years, there has been increasing interest in applying the idea of algebraic code superposition, also called “network coding” to the cooperative communications scenario. The network coding approach provides an efficient way to generate spatial diversity under the constraint of limited resources. One challenge is the problem of decoder design which should be able to cope with the complicated decoding situation at the destination . The work of proposed the convolutional code method to decode algebraically superposed code words, where a 64-state convolutional code is used for the XORed codeword composed of two 8-state encoders at the destination node. In , a code superposition scheme employing low-density generator matrix (LDGM) codes is proposed to reduce the decoding complexity at the destination. But in order to do the graph-based decoding, the systematic bits must be retained without superposition which means that the potential superposition diversity is lower than that obtainable from fully superposed codewords. In a combined low-density parity-check (LDPC) code construction scheme including two channel code components and one network code component is produced by random parity-check matrix generation under certain constraints. The network codes are actually the parity checks for two channel codewords; this necessitates more complicated relay operations than simple superposition. In this work, we consider the two-source one-relay cooperative model where the source-destination transmission and relay-destination transmission use a fading multiple access channel (MAC). We propose a cooperative coding scheme which is different from the previous work of where superposed codewords experience a channel orthogonal to that of the original transmission, and also different from the previous work of where simple codeword retransmission is employed in the multiple access Gaussian relay channel. Our scheme allows continuous transmission of superposed codewords by the relay and at the same time targets the challenge of coping with the interference introduced by the multiple access channel, thus making efficient use of communication resources to leverage spatial diversity gains. A corresponding sliding-window factor graph based decoding algorithm is used at the destination node to extract available spatial diversity gains. Computationally efficiency of the decoding algorithm is achieved through separation of the relevant SISO decoder modules and efficient connectivity via factor nodes corresponding to code and signal superposition operations. For convolutional codes, this separation affords a complexity advantage over decoding of the “nested code” ; for LDPC codes it affords a more efficient Tanner graph schedule than fully parallel decoding .