Cooperative anchor-less localization for large dynamic networks

High-definition location information is a key enabler for the emerging set of high-definition situation-aware (HDSA) systems. Applications such as search and rescue, blue force tracking, and exploratory rover localization often require rapid system deployment in unknown environments, in which case the setup of any pre-existing and fixed infrastructure becomes prohibitive. Further, such systems must often operate in harsh propagation environments such as inside buildings, in urban canyons, in caves, or on Martian terrain. In such infrastructure-limited scenarios, cooperation among network entities has the potential to provide the required self-location information for HDSA systems. In this paper we present a cooperative localization algorithm for use when there is no fixed localization infrastructure available. We evaluate the algorithm for large-scale networks with varied node mobility and then quantify performance gains when compared to classic cooperative techniques. A model for the communication between nodes is used that was obtained through an extensive measurement campaign around the MIT campus, using FCC-compliant UWB radios.