In this paper, the diversity multiplexing tradeoff (DMT) analysis of the non-coherent block-fading multiple antenna channel which uses a training-based channel estimation scheme at asymptotically high signal-to-noise ratios (SNRs) is extended to finite SNRs. This extension is performed for a single input multiple output (SIMO) maximal ratio combining (MRC) scheme. This analysis at finite SNRs is more useful because in practice, the training schemes operate at finite SNRs and their impact on DMT is more relevant at such SNRs. We show the non-applicability of the asymptotically high SNR relation, given by Zheng, to finite SNRs. We also show the equivalence of two existing training-based channel estimation schemes for any SIMO system, and using one of these, we compute the achievable finite-SNR DMT of the non-coherent SIMO-MRC scheme for two modes of the training scheme. We analyze the achievable finite-SNR DMT for different durations of training, modes of the training scheme, and SNRs. We show that the impact of the mode of the training scheme on finite-SNR DMT decreases as SNR increases. We also show that at asymptotically high SNRs, the achievable DMT in both modes of the SIMO-MRC scheme is equal to that of the non-coherent SIMO channel, as derived by Zheng.

The analytical derivation of the diversity-multiplexing tradeoff (DMT) for a half-duplex dynamic decode and forward (DDF) MIMO relay protocol has been regarded as an open problem. Recently, however, a minimization problem setting has been found, the solution of which corresponds to the DMT function for a half-duplex DDF MIMO relay protocol. In this paper, the DMT functions for three special half-duplex DDF MIMO relay protocols using two antennas at two of three nodes, source, relay, and destination nodes, and a single antenna at the other node are derived first. Then, the DMT function for a special half-duplex DDF MIMO relay protocol using two antennas at every node is derived. These DDF MIMO relay protocols are compared with one another and with some NAF MIMO relay protocols by simulation.

Recently, there has been growing interest in the design of wireless cooperative protocol to achieve higher diversity-multiplexing tradeoff among single antenna devices. We propose an automatic request for cooperation (ARC) scheme for wireless networks which can achieve higher order diversity by selecting the best relay. In this scheme, a source transmits a data packet towards a destination and a group of relays. The destination tries to decode the information from the source and if the detection is correct the process will stop. Otherwise, the destination transmits an ARC towards the relays. We utilize this ARC signal for selecting the best relay from the set of relays that have successfully decoded the source packet. The selected relay generates and transmits redundant information for the source packet. The destination combines the two packets received from the source and the best relay to improve the reliability of the packet. We analyze the packet error rate, spectral efficiency and diversity-multiplexing tradeoff of our proposal and compare them with some existing protocols. Analysis shows that our proposal can achieve higher diversity multiplexing tradeoff than conventional cooperative protocols.