Motion Parameter Capturing of Multiple Mobile Targets in Robotic Sensor Networks

A motion parameter capturing method is proposed to jointly estimate the initial positions and velocities of multiple mobile targets in robotic sensor networks. By using the time of arrival (TOA) measurements between the sensor nodes, the proposed method does not require any motion sensors. Noncooperative unconstrained linear least square (ULLS), constrained linear least square (CLLS), semidefinite programming (SDP), and mixed second-order cone semidefinite programming (SOCSDP) algorithms are designed by only exploiting the TOA measurements between anchor nodes and mobile targets. Then, the SDP and SOCSDP algorithms are extended to the cooperative approach in which the measurements between the mobile targets are also employed into the optimization model, as well as the measurements in the noncooperative approach. The simulations and real experiments show that the cooperative SDP and SOCSDP algorithms provide better performance than the noncooperative ULLS, CLLS, SDP, and SOCSDP. Compared with the linear ULLS and CLLS, the computation complexity of the convex SDP and SOCSDP is higher for a large number of variables and equality constraints. The positioning error of the SOCSDP is approximately identical to the SDP, but the SOCSDP runs faster than the SDP.