Opportunistic cooperative retransmission enhancements for the IEEE 802.11 wireless MAC

Publication Type:
Issue Date:
Full metadata record
Files in This Item:
Filename Description Size
Thumbnail01front.pdf117.32 kB
Adobe PDF
Thumbnail02whole.pdf1.16 MB
Adobe PDF
Cooperative retransmission protocols improve wireless transmission reliability by providing distributed channel diversity. Unfortunately, this diversity comes at the cost of increased protocol complexity and processing overhead, which limits the scalability of cooperative protocols in large networks. This Thesis introduces DAFMAC, an opportunistic retransmission protocol which operates without any explicit control messaging. DAFMAC uses passive transmission observations to select a suitable opportunistic relay using a distributed algorithm. The immediate benefit of reducing overhead is to enable a greater proportion of channel time available for data transmissions. The DAFMAC retransmission algorithm is compared to contemporary protocols through extensive simulations to evaluate network performance. The key result is that DAFMAC is able to meet or exceed the performance improvements of “coordinated” retransmission algorithms such as PRO and Δ-MAC, for metrics which include total network throughput, individual link fairness, energy efficiency, end-to-end transmission time and jitter. A proof-of-concept implementation of DAFMAC was deployed in a physical test-bed and was shown to significantly improve throughput in high path-loss links. This Thesis also derives a general retransmission model which is applicable to many distributed cooperative algorithms. Due to the complexity of implementing cooperative protocols in simulators, algorithms are typically only compared to traditional non-cooperative ARQ retransmissions. Further, analytic models typically include naїve simplifications that makes meaningful comparisons between algorithms impossible. The analytic model presented in this Thesis calculates the opportunistic retransmission outcome probability and includes detailed failure-mode results which may be used to rapidly compare algorithm performance with different configuration parameters in addition to comparing different protocols. Using the straightforward design principles of a retransmission algorithm, the model is able to accurately reproduce the cooperative performance results of a full state-based simulation. The retransmission model is independent of the channel model to facilitate performance analysis in different scenarios.
Please use this identifier to cite or link to this item: