公平的有向传感器网络方向优化和节点调度算法

Equitable Direction Optimizing and Node Scheduling for Coverage in Directional Sensor Networks

为了解决有向传感器网络中点目标覆盖控制问题,分别提出了两种方向优化算法和一个节点调度协议:改进的贪婪(enhanced greedy algorithm,简称EGA)、公平的方向优化(equitable direction optimization,简称EDO)算法和邻居节点调度协议(neighbors sensing scheduling,NSS).EGA基于覆盖最多未覆盖的目标数选取工作方向,其不足是可能忽略临界目标.EDO优化算法调节节点的工作方向,优先覆盖临界目标,公平分配感知资源,减小目标覆盖度的差异,EDO算法使用效用值评价每个方向对网络覆盖质量的贡献大小,影响效用值的因素包括每个方向上的目标数、目标的覆盖度和邻居节点的方向决策,EDO总是选择效用值最大的方向作为工作方向.NSS协议引入局部覆盖集的概念,通过局部覆盖集判断当前节点是否为冗余节点,并在考虑节点剩余能量时决定节点是否可以转为睡眠,调度协议允许一个节点加入多个覆盖集,覆盖集轮流工作,使网络生存期最大化.仿真实验结果表明,分布式的EDO算法比EGA算法具有更好的方向优化性能,临界目标的覆盖质量提高了30%,同时明显地提高了网络生存期.

To meet the coverage challenges arising in directional wireless sensor networks, this paper presents two distributed direction optimizing algorithms and a node scheduling： enhanced greedy algorithm （EGA）, equitable direction optimization （EDO） and neighbors sensing scheduling （NSS） protocol. EGA algorithm optimizes direction merely according to the amount of uncovered targets. It is used as the baseline for comparison. EDO adjusts the directions of nodes to cover the critical targets superiorly and allocates sensing resource among nodes fairly to minimize the coverage differences between nodes. The utility function is introduced in EDO to assess the value of a direction contributed to overall networks sensing. The factors which affecting the utility are composed of the targets in per direction, the coverage of targets and the neighbor＇s decision of direction. EDO always selects the direction with the maximum utility as the working direction. NSS arranges all sensors into multiple cover sets and allows a node to join several cover sets. Through employing local cover set, NSS identifies a redundant node and decides whether it can sleep while taking residual energy to account. Nodes are activated in turn and the energy is consumed evenly to prolong the network life. The simulation shows that EDO outperforms EGA up to 30% in terms of critical coverage, and the combination of EDO and NSS prolongs the lifetime distinctly.