| dc.description.abstract | Wireless ad hoc networks have a number of advantages over traditional, infrastructure-based networks. Robustness and easy deployment are two of the main advantages. However, the distributed nature of such networks raises a number of design challenges, especially when energy-efficiency and QoS requirements are to be taken into consideration. These challenges can only be met by allowing closer cooperation and mutual adaptation between the protocol layers, referred to as a cross-layer design paradigm. In energy-constrained wireless ad hoc networks, each node can only transmit to a limited number of other nodes directly. Hence, in order to reach distant destinations, intermediate nodes must relay the traffic of their peer nodes, resulting in multihop routes. The total energy consumption associated with a end-to-end transmission over such a route can be significantly reduced if the nodes are correctly configured. A cross-layer, optimization scheme, based on adaptive modulation and power control, is proposed in this thesis. The optimization scheme assumes that an existing route has been found, and allows QoS requirements in terms of end-to-end bit error rate and delay. Both transmission and circuit energy consumption is taken into consideration. By jointly optimizing all nodes throughout the route, the total energy consumption can be reduced by more than 50%, compared to a fixed-rate system. The adaptive system also exhibits superior capabilities to meet stringent QoS requirements. Results for both continuous and discrete rate adaptation is produced, and it is found that discrete adaptation causes only a small performance degradation, compared to the optimal, continuous case. Simulations also show that the system is vulnerable to inaccurate link state information. Finally, the effects of maximum-rate limitation and ignoring the circuit power consumption is investigated. | nb_NO |
