Robust Multi-Target Localization in ISAC Systems: Leveraging Multidimensional Scaling

Localization for multi-target systems (MTS) in anchor-free scenarios is receiving increasing attention, particularly for MTS, such as the Internet of Vehicles and swarms. This paper proposes a state-of-the-art fusion model based on the odometer and ranging measurements to address the anchor-free localization problem for MTS. Regarding varying measurements as a topological structure of the position, we formulate the optimization problem as a Procrustes problem that fully utilizes measurement information. Singular value decomposition is used to obtain the closed-form solution to the problem, which places minimal demands on computational resources. We also introduce an adaptive mechanism with a trade-off factor to enhance the model's flexibility. Additionally, the model applies to 2D and 3D cases and is generally suitable for most odometer and ranging technologies. Simulation experiments demonstrate that the positioning accuracy of the proposed scheme significantly improves over conventional optimization-based methods and outperforms the state-of-the-art.