A localization scheme based on Improving Dynamic Population Monte Carlo Localization method for large-scale mobile wireless aquaculture sensor networks

Abstract

Localization is one of the essential problems in wireless sensor applications (WSNs). Most range-free localization schemes for mobile WSNs are based on the Sequential Monte Carlo (SMC) algorithm. Multiple iterations, sample impoverishment and less sample diversity, leading to low localizing efficiency, are the most usual problems demanding to be solved in these SMC-based methods. An improved localization scheme for mobile aquaculture WSNs based on the Improving Dynamic Population Monte Carlo Localization (I-DPMCL) method is proposed. A population of probability density functions is proposed to approximate the unknown location distribution based on a set of observations through an iterative mixture importance sampling procedure, accompanied by node dynamic behaviours being analysed quantitatively or definitely. Threefold constrain rules are put forward in the I-DPMCL scheme to decrease the iteration number and trade off iteration number and enough valid samples to obtain the optimum iteration number. Then, these localization behaviours, especial delay, are predicted based on the statistical point of view. Moreover, performance comparisons of I-DPMCL with other SMC-based schemes are also proposed. Simulation results show that delay of I-DPMCL has some superiority to those of other schemes, and accuracy and energy consumption are improved in some cases of lower mobile velocity.