On Design Criterions For Distributed Space Time Codes And Relay Selection Schemes

It is well known that the cooperative diversity technology can greatly increase the reliability of a wireless communication system. This paper mainly focus on distributed space time codes and relay selection schemes which can achieve full diversity in a wireless relay network with a single antenna in the transmitter, in the receiver, in the relays, respectively.Firstly, the design criteria for distributed diagonal space time codes and multi-layer space time codes are analyzed in this paper. In the existing works, space time codes used in a wireless relay network are limited on unitary space time codes or orthogonal (or quasi-orthogonal) space time codes. However, for a general algebraic space time code, for example, a diagonal space time code, perfect space time code or a more general multi-layer space time code, the used matrixes do not limited on unitary. Because these space time codes have been proved to be able to achieve full diversity and good performance when applied in a MIMO system, a natural question is that how to apply a general space time code into a wireless relay network. This paper is on this direction. At first, a diagonal space-time code is applied into the wireless relay network. Then, a general multi-layer space time code generated by algebraic extension is also applied into the wireless relay network. The maximum-likelihood (ML) decoding is provided. Moreover, the optimal power allocations which can maximize the signal-to-noise ratio (SNR) and the Chernoff upper bounds of the pair-wise error probability (PEP) based on the ML decoding are calculated for these two kinds of space time codes respectively. Based on these bounds, two design criterions:the rank criterion and the singular value criterion are provided respectively. It is interesting that the singular value criterion obtained here is slightly different from the determinant criterion given in the paper [11] where the space time codes are unitary.Secondly, the relay selection schemes in a wireless relay network are analyzed in this paper. Assume that a transmitter has no channel information and a receiver has complete information of channels in a wireless relay network. First, by adding the assumption that each relay knows its own channels both from the transmitter to itself and from itself to the destination, the PEP of the multiple relay selection scheme based on relay ordering method proposed by Y. Jing and H. Jafarkhani in the paper[51] is calculated, when two relays are selected for relaying. Second, under an assumption that each relay has no any channel information, relay selection schemes based on relay ordering method are proposed. Furthermore, the PEP of the single relay selection and the double-relay selection are calculated, respectively. Finally, the simulation results are given. Both the theory and the simulation results show that the PEP of the multiple relay selection is lower than one of the single relay selection under the first assumption。Surprisingly, under the second assumption, the PEP of selecting two relays is higher than one of the single relay selection.Lastly, a joint scheme of relay selection and space time codes in wireless relay networks is introduced in this paper. Under an assumption that each relay has no any channel information in a wireless relay network, a scheme combining space time code with the relay selection is proposed in order to improve the performance of the system. It is proved theoretically that the Chernoff bound of the PEP of this combination scheme is much lower than one without applying Alamouti orthogonal space time code when two relays are selected. Moreover, when the signals transmitted are chosen from a QAM constellation and six relays in the network, the simulation result indicates that the gain of this scheme is more than 10dB at symbol error rate lower than 10-3.