Adhoc Wireless Network Control: Energy Efficiency and Hidden Terminal Considerations

摘要

We investigated challenges of wireless networks design with focus on energy management on battery dependent devices and the impact of a new hidden terminal problem on the operational efficiency of wireless networks. In the first part of this work we consider the optimal control of wireless networks operating with rechargeable batteries under general arrival, channel and recharge processes. The objective is to maximize total system utility while satisfying energy/power constraints. Starting from a downlink scenario, we propose a policy with decoupled admission control and power allocation decisions that achieves asymptotic optimality for sufficiently large battery capacity. Extensions to single-hop and multihop networks are also presented. Such policies are particularly suitable for satellite downlinks or sensor networks. In the second part of this work, we investigated performance unfairness (in terms of throughput and delay) on the Medium Access Control (MAC) mechanism of 802.11 standard in the presence of hidden terminals. To eliminate the impacts from the physical layer (PHY), we assume good channel. In this case, we may intuitively expect that all stations have equal success probabilities in their transmissions. This intuition is from the random access scheme of Carrier Sense Multiple Access with Collision Avoidance (CSMA/CA) which is employed by 802.11 MAC protocol. However, we will show that the intuition is not true due to hidden terminals. Instead, the following fact is revealed in this work: The success probability of a transmission is location- dependent. More specifically, the nodes far from the access point see more hidden terminals than those close to the access point, so they experience more collisions and thus smaller success probabilities. We build a model to analyze the throughput and collision probability for nodes at different locations, validate them via simulations and compare with the measurement results from experiments.