基于裁决门限辨析的无线传感网信号判定算法

Signal decision algorithm for wireless sensor networks based on decision threshold discrimination

为解决当前无线传感网节点识别算法中因信号强度低等因素导致难以有效识别恶意节点信号,且识别收敛性差,使其准确度不高的不足,提出基于裁决门限辨析机制的无线传感网sink抽样特征信号判定算法。定义信号接收模型,实现对sink节点接收信号的复原,还原对应的节点特征信号函数;利用正常信号分量的正交特性,采取复数域共轭处理的方式从正常信号解析形式中提取出恶意节点信号分量,对该分量的进行分组处理,提高该分量在背景环境中的强度;从分组中抽取最强的信号作为建立裁决门限的样本,当特征信号较弱的节点(恶意节点)触发裁决门限时,将直接通过节点特征信号函数进行特征信号频率还原,提高网络对节点特征信号的识别率。仿真结果表明,同AGDAM算法、AGDAM_Plus算法相比,所提算法能够有效降低恶意信号筛选过程中的信噪比,提高恶意节点正确筛选概率。

The signal decision algorithm with sink sampling characteristics for wireless sensor network based on decision thresh-old discrimination was proposed, to solve the problems such that it is difficult to effectively identify malicious node signal and the identification has poor convergence and low degree of accuracy, because of the low signal strength in the node identification algo-rithm for wireless sensor network at present. Signal reception model was defined, and sink node receipt signal and corresponding signal function with node characteristics were restored Orthogonal property of normal signal component was used, and signal component of malicious node was extracted from normal signal analytical form through plurality or conjugation and the compo-nent was grouped while its strength was improved in background environment. The strongest signal was extracted from the groups as the sample to build decision threshold. When nodes （malicious nodes） with weak characteristic signal triggered the de-cision threshold, the frequency of characteristic signal was restored through function of node characteristic signal, to improve the recognition rate of network for node characteristic signal. The simulation results indicate, compared with AGDAM algorithm and AGDAM _ Plus algorithm, the algorithm mentioned above can effectively reduce signal to noise ratio in screening malicious nodes and improve the probability in correctly screening malicious nodes.