Optimized low complexity sensor node positioning in wireless sensor networks

Localization of sensor nodes in wireless sensor networks (WSNs) promotes many new applications. A longer life time is imperative for WSNs, this requirement constrains the energy consumption and computation power of the nodes. To locate sensors at a low cost, the received signal strength (RSS)-based localization is favored by many researchers. RSS positioning does not require any additional hardware on the sensors and does not consume extra power. A low complexity solution to RSS localization is the linear least squares (LLS) method. In this paper, we analyze and improve the performance of this technique. First, a weighted least squares (WLS) algorithm is proposed, which considerably improves the location estimation accuracy. Second, reference anchor optimization using a technique based on the minimization of the theoretical mean square error is also proposed to further improve performance of LLS and WLS algorithms. Finally, to realistically bound the performance of any unbiased RSS location estimator based on the linear model, the linear Cramer–Rao bound (CRB) is derived. It is shown via simulations that employment of the optimal reference anchor selection technique considerably improves system performance, while the WLS algorithm pushes the estimation performance closer to the linear CRB. Finally, it is also shown that the linear CRB has larger error than the exact CRB, which is the expected outcome.