|
Author(s): |
Han Wang, Department of Computer Engineering, PLAUST, Nanjing, China
Wangdong Qi, Department of Computer Engineering, PLAUST, Nanjing, China |
|
Abstract: |
Centroid localization algorithm(CLA) is a distributed range-free localization method, with the advantages such as low-cost and simple to implement, whereas its localization error is big. APIT is an improvement of CLA .Location estimation of APIT is precise when anchors are deployed uniformly and densely whereas its computational cost is very high, moreover, the performance of APIT depends highly on the success of point in triangle test. When anchors and unknown nodes are deployed randomly, if some unknown nodes can not be covered by any anchor triangle, APIT will not perform on those nodes which make the average localization precision worse. In this paper, we design an improved centroid localization algorithm based on received signal strength. The core idea of our algorithm is, unknown nodes sort neighbor anchors according to corresponding RSS value, the estimation of location is then calculated as the centroid of the three closest anchors which satisfy unknown node in triangle test. A large number of simulations show that, when wireless sensor network is deployed randomly, the localization precision of RSSC is better than APIT; when WSN is deployed uniformly, the localization precision of RSSC is a little trivial. On average, computational cost of RSSC is less than a quarter of APIT.
|