WSN Node Localization Regularization Algorithm Based on Quasi Optimal Criterion Parameter Selection

Node localization technology is one of the basic research fields in Wireless Sensor Networks (WSN) applications. The coordinates of the unknown nodes can be determined by the Least Square Estimate (LSE), which is commonly employed in the WSN node localization. Due to the influence of multi-path fading, the distance can be obtained from the Received Signal Strength Indicator (RSSI). But in the experiments and applications, it is found that different spatial positions of the anchor nodes and the errors of the distance measurement sometimes lead to large location errors, which is called ill-posed problem. To solve this problem, condition numbers are selected to diagnose the ill-posed degree. When the ill-posed degree is weak, the LSE method can be utilized to localize. When the ill-posed degree is serious, the ridge estimate method is proposed to weaken the ill-posed problem, and quasi optimal criterion is proposed to choose the regularization parameter. Test results indicate that the ridge estimate method dilutes the influence of ill-posed problem obviously, and the location errors can be reduced to around 4 meters. Copyright © 2013 IFSA.