Cooperative Maximum Likelihood Target Position Estimation for MIMO-ISAC Networks

TL;DR

This paper proposes a cooperative ML-based framework for target position estimation in MIMO-OTFS integrated sensing and communication networks, demonstrating improved accuracy with more cooperating base stations.

Contribution

It introduces a novel cooperative ML approach that directly estimates target position in a common reference system, enhancing accuracy over traditional local estimation methods.

Findings

Position RMSE decreases with more cooperating BSs

Framework approaches the Cramér-Rao lower bound

Significant accuracy improvement over non-cooperative methods

Abstract

This letter investigates target position estimation in integrated sensing and communication networks composed of multiple cooperating monostatic base stations (BSs). Each BS employs a MIMO-orthogonal time-frequency space (OTFS) scheme, enabling the coexistence of communication and sensing. A general cooperative maximum likelihood (ML) framework is derived, directly estimating the target position in a common reference system rather than relying on local range and angle estimates at each BS. Positioning accuracy is evaluated in single-target scenarios by varying the number of collaborating BSs, using root mean square error (RMSE), and comparing against the square root of the Cram\'er-Rao lower bound. Numerical results demonstrate that the ML framework significantly reduces the position RMSE as the number of cooperating BSs increases.