Multitarget Passive Localization Model and Optimization in UHF RFID Scenarios

With the development of smart environments, device-free localization (DFL) has become a hot topic in the field of positioning. Even without users carrying networked devices, the environment can respond to their movements, enabling human-computer interactions and providing intelligent services. This paper proposes a sparse modeling and optimization localization algorithm based on passive ultra-high frequency radio frequency identification (UHF RFID) tags. The algorithm integrates target match value obtained from received signal strength (RSS) measurements with the overall impact of multiple targets on the environment, leveraging the sparsity of target positions relative to local maximum characteristic values to solve for the position of multiple targets. This paper establishes a feature information matrix that stores link characteristic information for each position index. The multi-target localization problem is modeled as an optimization problem under a sparse model. By analyzing the sparse solution, the sparsity of the solution vector is adjusted to reduce the impact of nonlinear measurement signals in multi-objective environments on the algorithm’s performance, thereby improving the accuracy of target counting and position estimation. In scenarios with an unknown number of targets, the average localization accuracy for 1 to 5 targets is 0.36 m, 0.69 m, 0.85 m, 1.14 m, and 1.31 m, respectively, and the average localization accuracy for dynamic targets is 0.49 m.