Radio resource allocation for multi-hop wireless networks using cross-layer optimization

In this thesis, we consider the problem of radio resource allocation for energy-efficient operation of multihop wireless networks, specifically for delay constrained traffic. The primary objective is to reduce energy consumption while providing end-to-end delay guarantee to the transported traffic across a wireless multihop network. We develop a framework to study the aforementioned problem and develop a suite of algorithms to optimize network performance.; We design and analyze a synchronous and distributed MAC protocol that provisions bandwidth for traffic over long time-horizons. Benefiting from its TDMA-based architecture and distributed reservation mechanism, the protocol guarantees sustained throughput at low delay even at high loads. We extend this problem by adding another optimization variable, namely signal power of transmission. Additionally, we incorporate the application's valuation of bandwidth or transmit power using a utility function. By explicitly accounting for the application's dependence on data rate and power, we are able to solve a network-wide optimization problem that maximizes the sum-utility of the network while providing data-rate guarantees per link.; Hitherto, the problem of traffic flow allocation was predetermined. In a multihop network, multiple routes are generally available between source and destination nodes. To achieve high rates it is important to use more than a single route. Determining the optimal flow allocation along each available path must take into account the underlying MAC layer dynamics. We consider this in the final part of the thesis. Specifically, we incorporate this degree of optimization and minimize the total average power consumption for end-to-end delay constrained traffic in a multihop network composed of links with time-varying channels. We develop a 2-tier hierarchical approach that adjusts flow on each route while adapting data transmissions at each link to the fade state of the wireless channel. Finally, we study the energy-delay trade-off for a wireless link in the presence of unforeseen errors due to imperfect channel side information. Lossless communications is achieved using a stop-and-wait ARQ approach. We design and implement a learning algorithm that "learns" the error process dynamically and adapts the packet transmission policy in a way that yields significant energy savings, albeit at higher average delay.