Distributed Cooperative Localization With Joint Estimation of Target Position and Transmit Power

Authors:
Publisher website
Google Scholar
Abstract
Localization in wireless networks is traditionally performed through centralized architectures that process measurements at a single node. However, such architectures not only lead to a high computational burden but also suffer from limited robustness, as the failure of the central node terminates the localization operation. They also require extremely high communication bandwidth, as all nodes must transmit their measurements to a central server. Moreover, assuming prior knowledge of each target node's transmit power is impractical, as it varies with battery level, hardware imperfections, and antenna orientation. To address these issues, this paper proposes a Distributed Cooperative Localization technique (DCLU), which jointly estimates the location and transmit power of the nodes. The joint maximum-likelihood formulation is non-convex and discontinuous, and cannot be solved using off-the-shelf optimization techniques. We address this challenge by constructing a convex approximation of the ML objective, which enables efficient distributed estimation using a coordinate descent algorithm. We further establish theoretical guarantees for convergence, including the rate of convergence. Compared with centralized approaches, DCLU offers enhanced scalability and robustness. We also derive the Cramer-Rao lower bound (CRLB) as a performance benchmark. Extensive simulations and on-field experiments demonstrate that DCLU achieves higher localization accuracy while significantly reducing computational complexity relative to existing techniques.
Funding Information
  • KAUST Office of Sponsored Research

This publication has 37 references indexed in Scilit: