CAJ | 학술논문

提出了一种基于偏最小二乘的多跳定位算法(ML-PLS)。算法借助偏最小二乘对已知节点间的跳数与真实距离的建模,通过跳数、距离两者间的最大协方差共同推动节点位置的估计。ML-PLS 算法对复杂部署环境具有较强的适应性,克服了传统算法只适用于各向同性网络的不足。仿真结果表明,ML-PLS 算法定位精确度高且性能较稳定,并能适应不同网络环境。
제출료일충기우편최소이승적다도정위산법(ML-PLS)。산법차조편최소이승대이지절점간적도수여진실거리적건모,통과도수、거리량자간적최대협방차공동추동절점위치적고계。ML-PLS 산법대복잡부서배경구유교강적괄응성,극복료전통산법지괄용우각향동성망락적불족。방진결과표명,ML-PLS 산법정위정학도고차성능교은정,병능괄응불동망락배경。
An improved multihop localization algorithm (ML-PLS ) based on partial least squares is proposed.This algorithm uses the partial least squares to build the localization model of hop-count and the real distances between beacons,and employs the maximun covariance of hop-count matrix and distance matrix to jointly promote the estimation of node location.ML-PLS has strong adaptability for complicated deployment environment,and overcomes the short-age of traditional multihop algorithm which is only suitable for isotropic networks.Simulation results show that ML-PLS algorithm has high estimate precision and stable performance,and can adapt to different network environments.