This paper is based on the method to find out global optimality of mean square error (MSE) for multi target tracking. In the previous work as probabilistic data association algorithm (PDA) is used to track each target separately, it is not possible to track multi target at a time while in joint probabilistic data association algorithm (JPDA) it is possible to track multi target randomly in a cluster. JPDA algorithm provides good tracking results in less dense cluster but for denser cluster it is less efficient as compare to previous algorithms (e.g. JPDA is in sub optimal in sense). The DAIRKF algorithm which is advanced than JPDA in the MSE sense for multi target tracking in high dense cluster but DAIRKF does not optimize MSE globally. The DAIRKF algorithm is simple in computation while PDA, JPDA algorithms provide exponential terms which increases computational complexity. DAIRKF algorithm involves the correlation between different targets and gives better optimality; it integrates random coefficient matrices for distributed multi targets which are not global in the sense of MSE optimization. Distributed Kalman filtering algorithm with integrated random coefficient matrices is used to achieve the global optimization of MSE. The real time VHDL simulation provide results to evaluate tracking performance by estimating MSE in DAIRKF and JPDA algorithms, both give some about complement results in high and low dense cluster with respect to each other. Using global optimal technique for optimal MSE of Integrated Random Coefficient Matrices Kalman Filtering provides better results than DAIRKF algorithm. The simulated result shows that the proposed algorithm is better than all previous algorithms (PDA, JPDA, and DAIRKF). It gives global MSE optimization which is efficient to track multi target precisely and accurately.

Full text: IJEIR_580_Final.pdf