Accurate and computation-efficient localization for mobile sensor networks

Localization is a fundamental problem in Wireless Sensor Networks (WSNs) and is more challenging than in the Static Sensor Networks in the result of location uncertainty caused by mobility of the sensor nodes. Existing range-free localization algorithms for WSNs are almost based on the sequential Monte Carlo Localization (MCL) algorithm. They either suffer from low sample efficiency or high communication- computation cost to achieve high localization accuracy. In this paper, we propose an accurate and computation efficient algorithm, called HCMCL, which could be categorized into MCL algorithms. In our algorithm, a series of distance constraint rules are created, which are based on the hop-count changes of sensor nodes. The constraints constructed from the rules can further reduce the size of sampling area and filter the samples more strictly. An existing technique called bounding-box is used in our algorithm to improve the sample efficiency by constructing a refined sampling area with the constraints constructed above. A method of weighting samples proposed in existing algorithm WMCL is also used in our algorithm to improve the localization accuracy. Simulation results show that the localization error in our proposed algorithm is much lower than the existing algorithms, especially when the sensor nodes move fast, and the computational cost is strongly reduced by a factor of up to 88 percent compared to WMCL.