MOO-SNLP: Multi Objectives Optimization for Sensor Network Localization Problem

This work highlights the key factors considered to evaluate the performance of a Localization System (LS) for W SNs. Indeed, the majority of the existing works concentrated essentially on the accuracy of the estimated location by the LS and ignored its performance in terms of localization cost, scalability and coverage. To address the accuracy/cost trade-off of a LS for WSNs, this work attempted to redefine and reformulate the Sensor Network Localization Problem (SNLP) as a Multi-Objective Optimization Sensor Network Localization Problem (MOO-SNLP). The latter requires the consideration of multiple objectives, like the position accuracy enhancement, with a reduced energy consumption, communication overhead and deployment cost of a localization approach. Moreover, we highlight the performances of the two most commonly-used localization approaches, Recursive Position Estimation (RPE) and Centroid Localization Algorithm (CLA), using different metrics under simulations. In fact, the obtained results show that the trade-off between location accuracy and implementation cost is promising in the area of multi-objective optimization related to SNLP.