In wireless sensor network (WSN) applications, node localization for complex system monitoring is a key problem to be
resolved. A new localization algorithm Based on Tikhonov regularization method is proposed for the ill-posed
multilateral localization problem, which includes building location model, selecting regularization parameter and
optimizing reference node number. In this algorithm, through converting ill-posed multilateral localization problem into
norm minimization problem, a feasible regularization matrix can be constructed and regularization parameter for
different situations can be determined using Morozov discrepancy criterion. The validity of the newly proposed method
is verified through experiments. Test results show that the location precision of the proposed algorithm is better than that
of the Maximum Likelihood Estimation (MLE) method, and the measurement error is about 1 meter while regularization
parameter &agr; is about 600 and the reference node number is 5.
|