Abstract: In view of the downhole conditions,the beacon node failure will exist a large number of blind area,this paper proposes a underground target track prediction algorithm based on markov chain model,the use of the existing wireless sensor networks and wireless radio frequency identification technology,combined with markov chain model to estimate target motion track,and the traditional centroid localization algorithm is improved,and adapt it to the prediction model,at the same time reduce the positioning error,improve the positioning coverage．Experimental results show:in hundreds of meters range error is smaller,which can realize accurate positioning.

Key words: Target track prediction,Markov chain,Wireless sensor network,Radio frequency identification,Centroid localization algorithm