Efficient relaying strategies for mitigating fading and interference
Rajendra Prasad Sirigina
Date of Issue2015
School of Computer Engineering
Centre for Multimedia and Network Technology
Relaying is considered as a key strategy to mitigate fading and interference in wireless systems. The interference mitigation in a relay-assisted two-user Z-channel is investigated in this work. The two-user Z-channel, where only one receiver is interference-limited, models the interference in heterogeneous cellular networks and ad-hoc wireless networks. In the Z-channel, the transmission strategies based on the probability of system outage may not be optimal for the interference-free user as the outage event is dominated by the interference-limited user. Instead, the diversity gain region (DGR) is an appropriate performance metric as it captures the trade-off between achievable diversity gains at both the users. The aim of this work is to analyze the DGR of the relay-assisted Z-channel. The achievable DGR of the full-duplex (FD) and half-duplex (HD) decode-and-forward (DF) relay-assisted Z-channels for both the single-user (SU) and multi-user (MU) encoding schemes are analyzed. Mitigation of the fading in a point-to-point channel with the HD DF relay is also investigated. Achievable DGR with both the SU codes and MU codes are presented for the direct transmission scheme with two antennas at the interfering transmitter. If the interference-free receiver is constrained to operate at the optimal diversity multiplexing tradeoff (DMT), the direct transmission scheme with MU codes is shown to achieve the same DGR as that of the single-user (SU) codes with the interference ignorant (II) detector. The receiver treats the interference as additive white Gaussian noise with the II detector. Then, the achievable DGR of the FD DF relay-assisted Z-channel with the SU codes is analyzed. It is shown that the FD DF relay clustered with the interfering source does not achieve the same DGR as that of the direct transmission scheme with two antennas at the interfering source. Hence, from the DGR perspective, it is better to have two co-located antennas at the interfering transmitter than a clustered FD DF relay. The present work studies whether the non-clustered FD DF relay achieves better DGR than the direct transmission scheme with single antenna at all the terminals. The achievable DGR of the MU codes based transmission schemes, namely FD partial decode-and-forward (PDF) and the simplified Han Kobayashi (SHK) schemes are investigated. Furthermore, the achievable DMT of the interference-limited receiver is optimized under the constraint that the interference-free receiver achieves the optimal DMT. It is observed that the FD PDF and the SHK schemes provide better DGR than that of the direct transmission scheme at lower interference levels only. This is because, as the interference level increases, the interfering transmitter should reduce the power of the private message to improve the achievable DMT at the interference-limited receiver. This reduction in the power may lead to the outage of the private message at the interference-free receiver for the higher interference levels. Closed-form expressions for the achievable DGRs with the half-duplex (HD) dynamic decode-and-forward (DDF) and HD selective DDF (SDDF) protocols are obtained. At the lower multiplexing gains, the HD DDF and the HD SDDF protocols are shown to achieve better DGR than that of the FD DF protocol. This is because, the interference-limited destination can use the joint decoder which is robust than the successive interference canceler (SIC). Next, the closed-form expressions for the DGR and the symbol error rate with SIC for the HD DF and HD amplify-and-forward (AF) relaying are derived. The minimum source-to-relay and relay-to-destination channel gains required to achieve the full diversity are also obtained. For a three terminal relay channel with an interference-free destination, selective link adaptive regenerative relaying (SLAR) is proposed. The SLAR scheme is shown to achieve full diversity while improving the power and spectral efficiency. In the SLAR scheme, the relay needs full instantaneous channel state information (CSI), which may need additional transmission bandwidth and power. SLAR with limited feedback (SLAR-LFB) protocol is proposed to reduce the costs associated with the CSI feedback. In SLAR-LFB, the relay needs only the average CSI, which incurs little additional cost. The SLAR-LFB protocol is also shown to achieve full diversity while improving the power and spectral efficiency.
DRNTU::Engineering::Computer science and engineering::Computer systems organization::Computer-communication networks