基于神经二部分裂结构的多智能体WSNs数据融合算法?

为进一步提高无线传感器网络WSNs( Wireless Sensor Networks)使用寿命，从提高算法数据融合效率角度出发，提出一种神经二部分裂结构的多智能体WSNs数据融合算法。首先，根据神经结构的主干与枝干承载信息量不同的原理，选取主干、枝干通讯链路并赋予较大能量，并给出主、辅中心节点选取方法；其次，设计了基于LMS的自适应加权融合算法，分别针对节点层级、枝干中心层级和主干中心层级进行逐层处理，实现了对神经二部分裂结构的数据融合；最后，通过与两种已有算法进行仿真对比，显示本文算法在Sink节点接收数据包，能耗等指标上均具有优势，验证了算法有效性。

NBMA-WSNs:Neural Bipartitions Structure Based Multiagent WSNs Fusion Algorithm

In order to further improve the wireless sensor network( WSNS) service life,from the angle of improving the efficiency of data fusion algorithm,,a neural bipartitions structure based multi-agent WSNs fusion algorithm was proposed. Firstly,according to the trunk and branches of the nerve structure bearing different amount of information, the trunk and branches were selected to construst the communication link,then the main,auxiliary center node’s se-lection method was also used. Secondly,the adaptive weighted fusion algorithm based on LMS was designed,which processed the node, the branches and the trunk center level to realize the aggregative for the neural bipartitions structure. Finally,through simulation compared with two existing algorithms,this algorithm shows the advantages at the indicators such as the Sink node’ss packet reception,the energy consumption,and so on,which verified the ef-fectiveness of the algorithm.