CS Paper Radar

Asynchronous Distributed Multi-Sensor Fusion Over Sensor Track Hypotheses

Ryan Draves, Nisar Ahmed, Eric Borcherding, Jonathan Chan

eess.SY ·submitted 2026-10-14 14:44 UTC · v1
published

Abstract

This work proposes a multi-target tracking solution that reasons over combinations of sensor-level tracks to perform distributed multiple hypothesis tracking (MHT) in an asynchronous multi-sensor network. This approach handles clustered target tracking and allows for efficient information pedigree summarization for distributed data fusion (DDF). By decomposing the association problem into partitioned hypothesis sets and fusing across sensor track combinations, the method reduces combinatorial complexity compared to traditional MHT approaches. Evaluation on a satellite constellation-based tracking scenario demonstrates improved estimation accuracy and data association quality compared to a reference target tracker with global nearest neighbor association.

Comment
Accepted to the 2026 IEEE International Conference on Multisensor Fusion and Integration (MFI), Plzeň. 6 pages, 6 figures

Closest papers

Nearest neighbours by meaning across the whole index.

Scalable Extended-Target Handover in Distributed Integrated Sensing and Communication

eess.SP ·2026-09-22 ·Liping Bai, Yu Ge, Ali Al Khansa, Hyowon Kim, Erik Leitinger +1 more
similarity 0.63
Abstract · 15 pages and 9 figures in total; includes 2 pages of supplementary material

Distributed integrated sensing and communication (DISAC) networks require multi-target tracking methods whose per-node computation and inter-base-station communication remain bounded as both the network and target populations grow. Existing multisensor fusion methods provide a strong theoretical foundation, but scalability is seldom treated as the primary design objective. We develop a grouped-measurement belief-propagation multi-target tracking (MTT) method and combine it with a proposed event-triggered target-handover protocol. When a tracked target is predicted to become observable at another base station, the owner transfers a predicted track message while retaining its local copy. Under bounded local workload and neighbor degree, we show that all variants of the proposed handover methods have network-size-independent per-node loads. Simulations demonstrate that handover reduces boundary track loss relative to uncoordinated processing and approaches coordinated-processing accuracy with lower communication overhead.

0.27

Information-Weighted Distributed Kalman Filter for Multi-Target Tracking over Switching Sensor Networks

eess.SY ·2026-10-09 ·Tirthankar Chakraborty, Arnab Maity
similarity 0.61
Abstract

Target tracking over wide areas using a network of sensor agents is a challenging problem when agents have limited sensing coverage and communication topologies vary over time. Existing distributed filtering approaches commonly assume that each agent, or at least one of its neighbors, can directly observe the target, and further rely on fixed network topologies, global network parameters, or multiple consensus rounds per time step. These requirements limit their applicability in wide-area network deployments. In this paper, we propose an Information-Weighted Distributed Kalman Filter (IWDKF) that relaxes these assumptions through a weighted-consensus mechanism applicable to arbitrary switching topologies satisfying joint connectivity. The proposed approach requires no prior knowledge of global network properties and limits communication overhead to a single consensus exchange per time step, while enabling information fusion even when some agents lack direct target measurements. The weighted-consensus covariance matrix is shown to remain positive definite, and the resulting consensus update is proven to maintain estimate consistency. It is also established that the proposed weighting strategy preserves the asymptotic agreement property of the underlying consensus protocol under joint connectivity. Numerical simulations of a multi-target tracking scenario involving a network of sensor agents demonstrate the improved estimation accuracy of the proposed IWDKF under both fixed and switching network topologies. The performance of the proposed filter is close to that of the centralized Kalman filter, and is compared with three existing distributed filters.

0.04

Distributed Target Tracking using Radar Networks

eess.SP eess.SY ·2026-08-19 ·Shao-Hsuan Hung, Raj Thilak Rajan
benchmarktheory similarity 0.60
Abstract

Distributed target tracking is essential for scalable and robust sensing systems, as it enables multiple radar nodes to cooperatively estimate a target state without relying on a centralized fusion center. In this paper, we present a fully distributed framework for single-target tracking in frequency-modulated continuous-wave (FMCW) radar networks. Each monostatic radar node observes local range and Doppler measurements and exchanges information only with neighboring nodes. Two consensus optimization-based estimators are developed. First, a Distributed Maximum A Posteriori (D-MAP) estimator is formulated for batch-based tracking, where prior state information is incorporated and the resulting optimization problem is solved using consensus-based alternating direction method of multipliers (ADMM). Second, a Distributed Extended Kalman Filter (D-EKF) is proposed for recursive tracking, where each node performs local prediction and correction followed by consensus ADMM-based information exchange. We derive the posterior Cramér-Rao lower bound (PCRLB) as a theoretical performance benchmark. Our simulation results show that D-MAP improves the accuracy of the estimation over the distributed maximum-likelihood baseline. These results demonstrate that the proposed framework provides a scalable and robust alternative to centralized radar tracking, given only local inter-node communication.

0.16

GenTrack3: Hybrid Stochastic-Deterministic Online Multi-Object Tracking with Cluster-Aware Association

cs.CV cs.RO ·2026-08-10 ·Toan Van Nguyen, Rasmus G. K. Christiansen, Dirk Kraft, Leon Bodenhagen
code/data similarity 0.59
Abstract · The content of this paper was included in the full manuscript of GenTrack family which has been submitted to the journal for possible public…

Multi-object tracking (MOT) involves maintaining consistent target identities as objects dynamically enter and leave a scene. Deterministic approaches, such as tracking-by-detection with data association, produce reproducible results and are computationally efficient, but they rely heavily on motion models and are sensitive to noisy detections that can lead to association errors. In contrast, stochastic methods explicitly model uncertainty and can better handle complex non-linear dynamics, albeit at the cost of increased computational complexity and variability arising from random sampling. This paper presents an online MOT framework that integrates deterministic and stochastic principles to achieve robust tracking under uncertainty. Furthermore, a novel track-to-detection matching approach is introduced to enhance scalability with increasing target numbers while supporting group tracking. The tracking inference mechanism employs a tracklet that includes identifiers, states, velocities, track penalties and track ages of targets, supporting a systematic tracking pipeline. Each target is associated with a stochastic particle set to compute the matching cost to detections. Reference implementations of the proposed approach and baseline trackers can be found on GitHub: https://github.com/SDU-VelKoTek/GenTrack3.

0.26

Distributed Observation Scheduling and Sensor Fusion For Large Agile Constellations

eess.SY ·2026-10-11 ·Nicholas Niziolek, Ryan Draves, Nolan Ales, Nolan Stevenson, Eli Jones +2 more
published similarity 0.59
Abstract · Presented at the 2026 AAS Rocky Mountain GN&C Conference, Breckenridge, CO (AAS 26-027). 19 pages, 16 figures

With the rapid increase in interest towards large scale earth observing satellite constellations, the problem of properly commanding and utilizing these systems has become increasingly relevant. This work focuses on a complete distributed approach to this problem, incorporating an observation task allocation and scheduling algorithm in addition to a sensor data fusion algorithm for end-to-end target coverage and processing. Observation task allocation and scheduling is performed in this work through the combination of the bid warped consensus based bundle algorithm (BWCBBA) and a lightweight observation scheduler for bid scoring. This method produces a unique mapping of tasks to command agents when allowed to iterate sufficiently. The approach to distributed data fusion in this work occurs primarily through an ownership model in which one satellite takes ownership of a system track in the constellation. After local information is shared between the sensing nodes to create the system tracks, the dedicated fusion nodes assume ownership for further information fusion and track information is disseminated throughout the constellation for data association. This paper analyzes the performance of these two algorithms over a set of increasingly decentralized architectures to highlight the performance losses compared to more traditional centralized methods.

0.59

Consensus-Based Distributed Filtering for Angles-Only Tracking in Cislunar Space

eess.SY cs.MA ·2026-10-22 ·Giovanni Romagnoli, Fabrizio Pilone, Giordana Bucchioni
publishedbenchmark similarity 0.57
Abstract · Accepted for presentation at IEEE NAVICON 2026

This paper investigates consensus-based distributed filtering strategies for multi-observer angles-only navigation in the cislunar space. The objective is to estimate the inertial state of a target spacecraft using only line-of-sight measurements collected by a network of observer spacecraft. This problem is particularly relevant for autonomous cislunar navigation, where ground support may be limited, and centralized fusion architectures may be problematic due to communication and computational constraints, and single-point failure risks. Three distributed filtering approaches are considered: the Distributed Extended Kalman Filter, the Information-Weighted Consensus Filter, and the Kalman Consensus Filter. All methods share the same local prediction and measurement-processing steps, but differ in the way information is exchanged and fused across the network. Their performance is evaluated and compared to a centralized Extended Kalman Filter, adopted as benchmark, and independent local filters in a high-fidelity cislunar scenario. Numerical simulations show that the distributed approaches significantly improve estimation accuracy with respect to independent filters. Monte Carlo sensitivity analyses are then conducted to characterize the robustness of the considered methods to finite consensus iterations and prior misalignment among the agents.

0.21

Rule-Sensitive Sensor Participation in Distributed Detection with Prescribed Boolean Fusion Rules

eess.SP cs.IT ·2026-11-11 ·Ahmad Tanha
similarity 0.56
Abstract

We study distributed binary detection when the fusion node must implement a prescribed Boolean decision rule. Under conditionally independent Bernoulli observations, we derive a signed-influence path identity valid for arbitrary Boolean rules; for monotone rules, it reduces to an exact Russo--Margulis representation in terms of coordinate influences. The identity yields rule-sensitive necessary conditions for a target Bayes error and an influence-weighted Kullback--Leibler feasibility bound. We then connect Boolean relevance to exact rule implementation and derive lower bounds on the number of sensors that must participate, including a heterogeneous bound based on the ordered per-sensor hypothesis separations. Numerical and Monte Carlo results illustrate both the tightening provided by the heterogeneous participation bound and the performance penalty that can arise when a prescribed fusion rule is retained instead of optimal likelihood-ratio fusion.

0.30

A Dynamic Generalized Kalman Consensus Filter for Switching Sensor Networks

eess.SY ·2026-10-01 ·Tirthankar Chakraborty, Arnab Maity
published similarity 0.55
Abstract · Added publication information and DOI on first page

Distributed state estimation is critical for applications such as surveillance, autonomous navigation, and wide-area monitoring, where sensor agents must cooperatively track targets using only local measurements and neighbor-to-neighbor communication. Existing distributed filters have been shown to achieve accurate estimation even under sparse inter-agent communication and limited sensing ranges. However, many of these methods rely on consensus parameters that depend on global properties of the communication graph, such as the maximum degree of the graph, and are therefore sensitive to changes in network topology. This limitation is particularly significant in sensor networks with mobile agents, where communication links change over time. This paper presents a Dynamic Generalized Kalman Consensus Filter for target tracking in sensor networks with switching communication topologies. The proposed algorithm computes information-based consensus weights using only locally available quantities, eliminating the need for global network parameters. Numerical simulations demonstrate that the proposed algorithm maintains estimation accuracy under switching network topologies and outperforms existing distributed filters in the given tracking problem.

0.50