Distributed solutions for rate control and maximum lifetime in wireless networks

This study focuses on fairness in wireless networks. Two fairness problems are addressed: end-to-end flow rate fairness in multihop wireless networks and lifetime fairness in wireless sensor networks.;In recent years, the advent of multihop wireless networks has greatly accelerated the research on bandwidth management in such networks to support new applications. While much research concentrates on the MAC layer, the users perception on these networks is however determined mainly based on the networks end-to-end effectiveness. It is important for us to develop flexible tools for traffic engineering in multihop wireless networks. In this study, two solutions are proposed to achieve end-to-end maxmin flow rate fairness in such networks.;A cross-layer design is firstly proposed for achieving end-to-end maxmin fairness in wireless mesh networks. In this approach, a generalized maxmin model is first proposed for multihop wireless networks. At the network layer, our design allocates network capacity to end-to-end flows for maxmin bandwidth allocation. At the MAC layer, our design achieves the allocated bandwidth shares for flows through a two-level weighted fair queuing algorithm. The proposed design is able to equalize the end-to-end bandwidth allocation to competing flows that share common bottlenecks, while fully utilizing the network capacity. Results of simulations are presented to demonstrate the effectiveness of the proposed solution in enhancing end-to-end fairness.;We also propose a fully distributed solution that is compatible with IEEE 802.11 DCF for achieving end-to-end maxmin fairness. We transform the global maxmin objective to four local conditions and prove that, if the four local conditions are satisfied in the whole network, then the global maxmin objective must be achieved. We then design a distributed rate adaptation protocol based on the four conditions. Whenever a local condition is tested false at a node, the node informs the sources of certain selected flows to adapt their rates such that the condition can be satisfied. Comparing with previous work, our protocol has a number of advantages. First, it does not modify the backoff scheme of IEEE 802.11. Second, it replaces per-flow queueing with per-destination queueing. Packets from all flows to the same destination is queued together. Third and most important, our protocol achieves far better fairness (or weighted fairness) among end-to-end flows than previous work.;Wireless sensor networks have a wide range of applications in habitat observation, seismic monitoring, battlefield sensing, etc. As another type of multihop wireless network, a sensor network consists of battery-powered sensor nodes that are limited in energy supply. An important problem of wireless sensor networks is maximizing the operational lifetime of a sensor network. The lifetime of a sensor network is defined as the lifetimes of all sensors that produce useful data. A centralized solution proposed by previous work requires solving a sequence of linear programming problems. The computation overhead can be prohibitively high for large sensor networks. Collecting the complete information about the network and uploading the complete forwarding policies to all nodes require significant amount of transmissions, particularly for nodes around the sink. We propose a fully distributed progressive algorithm which iteratively produces a series of lifetime vectors, each better than the previous one. Instead of giving the optimal result in one shot after lengthy computation, the proposed distributed algorithm has a result at any time, and the more time spent gives the better result. We show that when the algorithm stabilizes, its result produces the maximum lifetime vector. Furthermore, the algorithm is able to converge rapidly towards the maximum lifetime vector with low overhead.