Distributed partial decoding in cooperative communication systems

Files in this item

Files Size Format View
3362340.PDF 5.736Mb application/pdf Thumbnail

Show full item record

Item Metadata

Title: Distributed partial decoding in cooperative communication systems
Author: Karkooti, Marjan
Advisor: Cavallaro, Joseph R.
Degree: Doctor of Philosophy thesis
Abstract: Increasing demand for wireless services is putting major pressure on network resources, which demands a new paradigm with a more efficient design. Recently, cooperative communications has emerged as a viable option for future wireless devices. Major improvements have been made in the theoretical analysis of cooperative communications and relay channels in recent years. But, most of the analyses have some simplifying assumptions that may not be valid in practice. These assumptions include using infinite-length block codes, zero processing delay, etc. This thesis considers cooperative communications from a practical point of view and identifies the benefits of cooperation when some of the theoretical assumptions are relaxed or changed. Several techniques are introduced to reduce the complexity of the system with minimal performance loss. We set up a framework for system design and show adaptability of our techniques to different scenarios. Our main focus is on the decode-and-forward relaying strategy with low density parity check (LDPC) codes. The thesis contributions in the field of cooperative communications fall into two main categories: algorithms and architectures. First, we focus on the complexity reduction in the algorithms and propose 'distributed partial decoding'. We demonstrate the benefits of partially decoding the codeword at the relay and distributing the decoding load between the relay and the destination. This results in major savings in terms of processing power and time at the relay with a very small loss in system performance. The architectural complexity and overhead of this scheme is much smaller than the original decode and forward strategy. The second contribution of this thesis is the design and implementation of a flexible LDPC decoder architecture that supports a family of LDPC with a variety of code rates and block lengths. This architecture is very suitable for cooperative environments where the cooperating pair and channel conditions and hence the code parameters are not known in advance. The third contribution relates to leveraging puncturing in cooperation. This work is the first to analyze cooperative communication with punctured LDPC codes. We propose structured puncturing patterns for quasi-cyclic LDPC codes and analyze the tradeoffs in designing good puncturing patterns for cooperative environments.
Citation: Karkooti, Marjan. (2009) "Distributed partial decoding in cooperative communication systems." Doctoral Thesis, Rice University.
Date: 2009

This item appears in the following Collection(s)