基于分簇和融合补偿策略的多维标度定位算法

Advanced MDS-MAP Localization Algorithm with Clustering and Fusion Compensation Strategy

针对经典的多维标度(MDS-MAP)定位算法在大规模无线传感器网络中存在的定位功耗大和精度低的问题,改进后的MDS-MAP算法将节点作为簇头时的剩余能量、能耗均衡性与局部密度的综合指标进行评估后再进行分簇,形成的簇具有良好的连接性与较低的能量损耗。针对部分不满足拼合规则的节点,提出了一种利用度量策略来获得节点间未知的欧氏距离的方法,并用角度判别法消除干扰解。在对公共节点进行补偿后,使用改进的规则进行簇间合并。仿真比较结果表明,提出的基于分簇与融合补偿策略的多维标度定位算法具有较低的拼合要求、高定位精度以及强鲁棒性,有利于拓展网络和降低定位功耗。

The classic multi-dimensional scaling positioning(MDS-MAP)algorithm has the problem of high energy consumption and low positioning accuracy in large-scale wireless sensor networks.The improved MDS-MAP algorithm evaluates the residual energy,the balance of energy consumption and the local density of the cluster head when nodes are used as cluster heads and then the clustering is performed.The clusters have good connectivity and low energy loss.To overcome the limitation of the flattening rule,this paper proposed a method to obtain the unknown Euclidean distance between nodes,and the method of angle discrimination was used for eliminating the solution of interference.After compensating for common nodes,the improved rules was applied to the inter-cluster merging.Simulation and comparison results indicate that the proposed advanced MDS-MAP localization algorithm with clustering and fusion compensation strategy has lower splitting requirements,high positioning accuracy and robustness,which is good for the network expansion and the reduction of positioning power consumption.