针对无线传感器网络中实时数据收集具有较高的延时问题, 提出了一种改进的无通信冲突的分布式数据聚集调度近似算法。该算法首先在最大独立集的基础上建立一棵根在sink的数据聚集树, 然后各个节点按数据聚集树分层进行数据调度。在数据聚集树的构造过程中, 对于两个相距两跳的支配点, 它们共同的、相距两跳的支配点, 通过距sink最近的支配点加入数据聚集树; 而在数据调度过程中, 采用一种新的选择标准从竞争集中选择节点进行数据调度。通过这两方面的改进, 有效地降低了数据的聚集延时。理论分析表明, 该算法的延时上界为14R+Δ; 仿真模拟的结果表明, 该算法产生的数据聚集延时远低于现有算法。
This paper presented an improved distributed data aggregation scheduling algorithm without communication collision for minimum data aggregation latency due to existing algorithms have high time latency for data collection in wireless sensor networks. In this algorithm, it constructed a data aggregation tree rooted at the sink firstly. And then, the node could be scheduled layer by layer according to the data aggregation tree. In the process of the constructing the data aggregation tree, for the common neighboring dominators of two adjacent dominators, it selected a dominator which was close to the sink to join the data aggregation tree. It used new criteria for node selection amongst available competitors in the data aggregation scheduling scheme. Using these modifications, it reduced the latency for the sink collecting all sensors' data effectively. The theoretical analysis shows that the algorithm has a latency bound with 14R+Δ. Simulation results show that this algorithm has lower average latency than previous works.