Many applications and protocols in wireless sensor networks need to know the locations of sensor nodes. A low-cost method to localize sensor nodes is to use received signal strength indication (RSSI) ranging technique together with the least-squares trilateration. However, the average localization error of this method is large due to the large ranging error of RSSI ranging technique. To reduce the average localization error, we propose a localization algorithm based on maximum a posteriori. This algorithm uses the Baye's formula to deduce the probability density of each sensor node's distribution in the target region from RSSI values. Then, each sensor node takes the point with the maximum probability density as its estimated location. Through simulation studies, we show that this algorithm outperforms the least-squares trilateration with respect to the average localization error.
|Number of pages||7|
|Journal||International Journal of Distributed Sensor Networks|
|Early online date||15 Dec 2011|
|Publication status||Published - 1 Jan 2012|
Bibliographical noteThis paper was supported by the National Natural Science Foundation of China (Grant no. 61003272, no. 61033009, no. 61170076, and no. 61103001), the Guangdong Natural Science Foundation (Grant no. 10351806001000000), and the Shenzhen Science and Technology Foundation (Grant no. JC201005280408A and JC2009D3120046A).
LU, K., XIANG, X., ZHANG, D., MAO, R., & FENG, Y. (2012). Localization algorithm based on maximum a posteriori in wireless sensor networks. International Journal of Distributed Sensor Networks, 8(1), 1-7. https://doi.org/10.1155/2012/260302