摘要：The positioning technology, as one of the keytechnologies of wireless sensor network, can make each nodeacquire its own position information in the network to achievetarget monitoring, data acquisition and behavior prediction.Under indoor environment, owing to the influence of obstacles,signal transmission between mobile nodes generally is underthe non-line-of-sight environment, causing more localizationerrors. In this paper, focused research on the localizationalgorithm in the initial state and mobile state when nodes enterthe network is conducted under non-line-of-sight environment.Firstly, a multi-signal-based localization algorithm is proposedagainst the initial localization stage at which mobile unknownenters the network. Appropriate anchor nodes are selected toparticipate in localization computing through the signalstrength value, thereby narrowing the area where unknownnodes locate and improving the localization accuracy of nodes.Meanwhile, aiming at the problem of non-line-of-sight errors,anchor nodes with low errors are selected by introducingnon-line-of-sight inhibitory factor to reduce the impact ofnon-line-of-sight measured value on the localization results.Secondly, aiming at the motion state of mobile nodes, alocalization algorithm based on historical localization isproposed. Mobile nodes are located according to the number ofanchor nodes around node localization moment in conjunctionwith the historical localization information of nodes.Meanwhile, the coordinate computing method is modified byintroducing a modification value to make the coordinate valuesof nodes closer to the real value. The simulation result indicatesthat the proposed localization method can improve thelocalization accuracy effectively and has good adaptability.
