基于无线传感器网络特性,提出一种不等簇半径非均匀分簇策略,根据各节点至Sink的距离,由近至远依次采用等比递增的不等簇半径进行分簇。并从理论上证明该策略可有效地均衡无线传感器网络的能量消耗。实验结果表明:距离Sink较近的节点,采用较小的簇半径成簇,使得簇头节点可保留更多能量进行簇间数据转发;距离Sink较远的节点,采用较大的簇半径成簇,使得簇头节点可保留更多能量覆盖更大范围的节点进行簇内数据收集,从而使得整个网络的负载比簇半径恒定的分簇网络更加均衡。
Based on the characteristics of wireless sensor networks, clustering with unequal cluster-radius was proposed. It is proved that according to the distance between sensors and the Sink from near to far, clustering with geometric increments cluster-radius can effectively balance the network energy consumption. Sensors near the Sink should be clustered with a smaller cluster-radius, which makes the headers of the cluster save more energy to relay data from outer headers. Sensors far from the Sink should be clustered with a bigger cluster-radius, and then the headers of the cluster can save more energies to cover more sensors, and thus can collect more data in the cluster. Consequently the whole network load can be more balanced than the network clustered with fixed cluster-radius.