Fusion-Based State Estimation for Localization and Synchronization of Distributed Radar Sensor Netwo
Abstract
The United States Navy faces several significant obstacles in the near future. Uncertain littoral environments, asymmetric engagemen,ts, the proliferation of unmanned (especially aerial) vehicles, high-speed and maneuverable hypersonic vehicles, and an ever-increas,es, the U.S. Navy must increase its reliance on networked radar operations with enhanced cooperation to maintain battlefield superio,rity and ensure timely, accurate, and assured intelligence, surveillance, and reconnaissance inputs for informed decision making. In, addition, these distributed radar nodes need autonomous and mobile capability, and the overall network must be resilient when deplo,ated from one another. They are superior in terms of system survivability as they do not have a single point of failure and provide,many advantages in terms of performance, such as increasing target localization accuracy, mitigating the blind velocities caused by,Doppler aliasing, providing three-dimensional velocity estimation, and improving the power gain and directivity of beamforming (BF),missions. Moreover, synthetic aperture radar (SAR) imaging modalities stand to improve by enabling quicker updating of scene images,, providing multi-look angle measurements of a scene from multiple angles, and increasing the synthetic aperture size.For all the ben,efits of a distributed radar network, many fundamental challenges stand in the way of its practical deployment. The primary challeng,e is synchronizing all the network sensors in time, frequency, and phase. A secondary challenge in BF and SAR applications is that t,he sensor nodes positions relative to each other and the scene of interest must be precisely known on a per-pulse basis. Although t,he Global Positioning System (GPS) is used for synchronization and positioning, GPS has a timing standard deviation of 15 ns and wil,l likely be unavailable in near-peer adversary situations; thus, these procedures will need to be performed independently. This yiel,ds a tertiary challenge of estimating the moving sensors positions as the estimations will depend on inertial measurement unit (IMU,) measurements, which diverge quickly from the true position.The PI proposes a solution using current state-of-the-art fusion method,ologies and novel esti- mation techniques to synchronize and localize moving radar sensors. Given the close relationship between the, time, phase, and frequency synchronization and the estimation of relative distances between nodes, time-of-flight (TOF) measurement,s can be utilized to validate the accuracy of position estimates made by an IMU. Because the position estimates are assumed to have,some error, determining an optimal position estimate, TOF between all nodes, and the phase, time, and frequency offsets between all,the network sensors is well-suited for a state estimator such as the Kalman or particle filter. Thus, this proposal will investigate, the fundamental mathematics, algorithm development, and system-of-systems architecture needed for distributed, mobile radar system,synchronization, localization, and navigation.The outcomes of this proposed research will enable coherent distributed radar operatio,n and sensor node localization even in situations where GPS is not available, giving such systems a dramatic advantage in hostile en,vironments. It would also estimate each node s position and the phase, time, and frequency offsets in each node by using measured ph,ase offsets, TOF between nodes, and the kinematic values provided by an IMU. When implemented, the proposed solution can enable auto,nomous sensor swarms cap,blic Release.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Aug 05, 2022
- Source ID
- N000142212614
Entities
People
- Jay W. McDaniel
Organizations
- Office of Naval Research
- United States Navy
- University of Oklahoma