可扩展的分布式集成感知与通信中的扩展目标交接
Scalable Extended-Target Handover in Distributed Integrated Sensing and Communication
浏览论文内容
中文总结 AI 辅助
针对分布式集成感知与通信网络，提出分组测量置信传播多目标跟踪与事件触发目标交接协议，实现与网络规模无关的每节点负载，降低通信开销并减少边界轨迹丢失。
中文摘要 AI 辅助
分布式集成感知与通信（DISAC）网络需要多目标跟踪方法，这些方法在网络和目标规模增长时，其每节点计算量和基站间通信量保持有界。现有的多传感器融合方法提供了坚实的理论基础，但可扩展性很少被视为首要设计目标。我们开发了一种分组测量置信传播多目标跟踪（MTT）方法，并将其与所提出的事件触发目标交接协议相结合。当被跟踪目标预计在另一个基站变得可观测时，拥有者会传输预测的轨迹消息，同时保留其本地副本。在有界本地工作负载和邻居度下，我们证明了所提出的交接方法的所有变体都具有与网络规模无关的每节点负载。仿真表明，与无协调处理相比，交接减少了边界轨迹丢失，并以较低的通信开销接近协调处理的精度。
英文摘要
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.
发表机构
- Chalmers University of Technology（查尔姆斯理工大学）
- Massachusetts Institute of Technology（麻省理工学院）
- Orange Labs（法国电信实验室）
- Chungnam National University（忠南国立大学）
- Graz University of Technology（格拉茨工业大学）
机构由 AI 辅助整理，请以论文原文为准。
