关键词: 无线传感网络, 定位, 到达时间, 核局部保持映射, 核主成分分析

Abstract: This paper aims at reducing the influence of ranging error on localization accuracy. A Wireless Sensor Network (WSN) localization algorithm was proposed, which was based on Kernel Locality Preserving Projection (KLPP). The algorithm took transmission time vector between nodes as input, and a model was established by KLPP, which can reflect partial information network topology structure. The simulation results indicate that compared with Kernel Principle Component Analysis (KPCA), the algorithm can achieve higher localization accuracy in Time of Arrival (TOA) localization and reduce the influence of ranging error on localization accuracy effectively in complex environment.

Key words: Wireless Sensor Network (WSN), localization, Time Of Arrival (TOA), Kernel Locality Preserving Projection (KLPP), Kernel Principle Component Analysis (KPCA)