Location estimation of sensor nodes using learning movement patterns

The definition of a localization system among sensor nodes is a fundamental issue for many applications of wireless sensor networks (WSNs). Because sensor networks may be deployed in inaccessible area the position of sensor nodes may not be predetermined. Thus, a localization system is required in order to provide position information to the nodes. The cost and limited energy resource associated with common, mobile nodes prohibits them from carrying relatively expensive devices such as Global Positioning System (GPS). This paper proposes a mechanism that allows non-GPS-equipped nodes in the network to derive their approximated locations from a limited number of GPS-equipped nodes. Non-GPS nodes estimate the distance by measuring the Received Signal Strength Indicator (RSSI).In our method, Hidden Markov Model (HMM) is used to estimate the approximated location of distance estimates with the help of mobility models. Simulation studies show that our solution is capable of producing good estimates-equal or better.