Dynamic networks of physical nodes that require self-localization and synchronization are on the horizon. Existing infrastructure-based positioning and timing solutions are not fully sufficient for many applications. A peer-to-peer architecture known as cooperative localization and synchronization (CLAS) is a promising alternative when infrastructure is unavailable or insufficiently accurate. Efficient solutions for small, low-cost devices and platforms will be created by blending the functions of communication, localization, and synchronization into a single paradigm. Before CLAS can be widely adopted, we believe that it is necessary to develop localization and synchronization methods that do not impose additional constraints on the wireless communication system and are able to leverage a wireless broadcast transmission modality for enhanced scalability. We propose the first CLAS data fusion algorithm capable of operating over an unstructured communication network with no requirement for a custom pairwise communication protocol or global time anchors. The algorithm utilizes transmit/receive timestamp pairs from any received message, including a broadcast, and outputs an updated joint estimate of spatial and temporal state parameters. Both centralized and distributed versions of the algorithm are shown to be accurate and robust when used in conjunction with a commercially available ultra-wideband (UWB) radio. Results are presented for both simulated timestamps and data collected with a broadcast network of UWB radios.

Cooperative localization is an enabling technology for the IoT that will introduce innovative services for modern convenience and public safety. Particle-based belief propagation (BP) is a state-of-the-art method for cooperative localization. However, in large and dense cooperative localization networks, particle-based BP suffers from particle degeneracy. In this paper, we propose a new method that combines particle-based BP and particle flow (PF) and can avoid this detrimental effect. To perform operations on the graph effectively, particles are moved towards regions of high likelihood based on the solution of a partial differential equation. We show that the proposed PF-BP algorithm can significantly outperform conventional particle-based BP in accuracy and runtime.

Location-aware networks enable new services and applications in fields such as autonomous driving, smart cities, and the Internet-of-Things. Network localization and navigation (NLN) is a recently proposed paradigm for accurate ubiquitous localization without the need for extensive infrastructure. In NLN, devices form an interconnected network for the purpose of cooperatively localizing one another. This paper introduces Peregrine, a system that combines real-time distributed NLN algorithms with ultra-wideband (UWB) sensing and communication. The Peregrine software integrates three NLN algorithms to jointly perform 3-D localization and network operation in a technology agnostic manner, leveraging both spatial and temporal cooperation. Peregrine hardware is composed of compact low-cost devices that comprise a microprocessor and a UWB radio. The contribution of each algorithmic component is characterized through indoor network experimentation. Results show that Peregrine is robust, scalable, and capable of sub-meter accuracy in challenging wireless environments.