Géolocalisation des DM : principes et applications

Abstract

Position location usually relies on direct observation to/from conventional landmarks with known positions from/to a Node of Interest (NOI). Nonetheless, in an Ad-Hoc Wireless Sensor Network (AHWSN), nodes are often unable to establish a direct connection with the available Access Points (APs). In such a scenario, neighboring nodes may supply cooperative information to enable inference of the location of a given NOI in a network. In this paper we examine the feasibility of relational techniques in multihop environments to estimate position location. Two novel position estimation techniques are presented: the Relative Proximity Algorithm (RPA) and the Enhanced Relative Proximity Algorithm (ERPA). RPA and ERPA can operate as range-based or as range-free techniques, which makes them both attractive and flexible solutions for position estimation in AHWSNs. The performance of these techniques is characterized and found to be related to the number of cooperating nodes, the number of APs available in the network, and the presence of measurement noise. RPA and ERPA are also compared to several known position location methods reported in the literature, and it is shown that they achieve adequate location estimation accuracy with some advantages in terms of the number of access points required and network traffic overhead.