Adaptive Equalization and Capacity Analysis for Amplify-and-Forward Relays
Thesis DisciplineElectrical Engineering
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
Recent research has shown that multiple-input multiple-output (MIMO) systems provide high spectral efficiencies and error performance gains. However, the use of multiple antennas in mobile terminals may not be very practical. Certainly there is limited space and other implementation issues which make this a challenging problem. Therefore, to harness the diversity gains afforded by MIMO transmitter diversity techniques, while maintaining a minimal number of antennas on each handset, cooperative diversity techniques have been proposed. In addition, attention has also been given to combining wireless relaying systems with MIMO techniques to improve capacity, coverage, and obtain better diversity at the expense of increased node complexity.
This thesis considers the design and analysis of cooperative diversity systems and MIMO amplify-and-forward relaying systems. In particular, we investigate adaptive time- and frequency-domain equalization techniques for cooperative diversity systems using space-time block codes (STBC). For MIMO relaying systems, we analyze the ergodic capacity of various systems and compare different amplify-and-forward methods in terms of system capacity performance.
We propose a new block time-domain adaptive equalization structure for time reversal-space time block coding (TR-STBC) systems, which eliminates the separate decoder and also the need for explicit channel state information (CSI) estimation at the receiver. Our simulation results show that the time-domain adaptive block equalizer performs better than the frequency-domain counterpart but at the cost of increased complexity. Then, we extend this time-domain adaptive equalization scheme to distributed TR-STBC systems. We also develop a frequency-domain counterpart for the distributed systems. Our simulation results show that the adaptive algorithms work well for Protocols I and III proposed by Nabar et al. The time-domain adaptive algorithms perform better than the frequency-domain algorithms, and overall the Protocol I receivers outperform the Protocol III receivers. We also show that, if only the Protocol III receiver is used, it can be susceptible to noise amplification due to a weaker source-to-relay link compared to the relay-to-destination link. This problem can be mitigated by using the Protocol I receivers with some extra complexity but much superior diversity performance.
We also present an ergodic capacity analysis of an amplify-and-forward (AF) MIMO two-hop system including the direct link and validate the analysis with simulations. We show that having the direct link improves the capacity due to diversity and quantify this improvement. We also present an ergodic capacity analysis of an AF MIMO two-hop, two relay system. Our results verify the capacity gain of relaying systems with two relays due to the extra diversity compared to a single relaying system. However, the results also show that when one of the source-to-relay links has a markedly higher SNR compared to the other, a single relay system has better capacity than a two relay system.
Finally, we compare three types of relay amplification methods: a) average amplification, b) instantaneous channel amplification, and c) instantaneous power amplification. The instantaneous power amplification method has a higher mean capacity but with a higher variance. Also, it requires additional information at the destination and would create enormous overheads compared to the other methods. We also find that the instantaneous channel amplification method has almost no advantage in terms of the mean capacity but its capacity is less variable than the average amplification method. On the other hand, the average amplification method is simpler to implement as it does not require channel estimation at the relaying terminal.