Cross-layer near-far effect mitigation for wireless ad hoc networks.

摘要

Ad hoc wireless networks provide networking capability in situations where no fixed infrastructure exists. In these networks, the communication between nodes can be either direct or via relaying nodes. While these networks provide useful networking capabilities in anomalous situations, their performance is strongly affected by the interference created within the network.;In this thesis, we consider the particular case of Code Division Multiple Access (CDMA) based ad hoc networks, which have been shown to have desirable properties, such as interference and jamming resistance, robustness to multipath fading, low probability of intercept and soft degradation of performance with increased network load. Despite of these important advantages, CDMA ad hoc networks are very sensitive to uneven interference, namely to the near-far effect. Traditional methods to mitigate the near-far effect, such as power control and multiuser detectors, prove to be ineffective in an ad hoc network setting due to the peer-to-peer transmission nature of the network on one hand, and due to the high implementation complexity of the multiuser detectors on the other hand.;In this dissertation, we present several techniques to mitigate near-far effect in wireless ad hoc networks at upper levels of the protocol stack. Our design emphasis is on using a cross-layer architecture and on enforcing cooperation among various network nodes.;Our proposed work is focused onto two main components: (a) at the physical/data link layer, proposing code-spread CDMA techniques for interference mitigation, (b) at the network layer, proposing interference aware, energy efficient routing algorithms, that promote near-far effect mitigation and cooperation among nodes.;Several algorithms are proposed for the network level interference management approach, using location information, near-far effect impact estimates for the potential relaying nodes and game theoretic models.;Our focus in this thesis is mainly on data networks, and we show significant performance improvements for our proposed algorithms, in terms of both energy efficiency and throughput gains. The penalty to pay is always an increase in the implementation complexity. The last chapter of our work is especially focused on quantifying such complexity/performance tradeoffs for ad hoc wireless networks. Effective throughput and energy gains are quantified via analysis and simulation, when the effect of increased overhead due to more complex algorithms is taken into account. We show that our proposed algorithms still exhibit significant performance gains, despite of a slight increase in implementation complexity and overhead.