Ali Onder Bozdogan

Improved assignment with ant colony optimization for multi-target tracking

Detecting and tracking ground targets is crucial in military intelligence in battlefield surveillance. Once targets have been detected, the system used can proceed to track them where tracking can be done using Ground Moving Target Indicator (GMTI) type indicators that can observe objects moving in the area of interest. However, when targets move close to each other in formation as a convoy, then the problem of assigning measurements to targets has to be addressed first, as it is an important step in target tracking. With the increasing computational power, it became possible to use more complex association logic in tracking algorithms. Although its optimal solution can be proved to be an NP hard problem, the multidimensional assignment enjoyed a renewed interest mostly due to Lagrangian relaxation approaches to its solution. Recently, it has been reported that randomized heuristic approaches surpassed the performance of Lagrangian relaxation algorithm especially in dense problems. In this paper, impelled from the success of randomized heuristic methods, we investigate a different stochastic approach, namely, the biologically inspired ant colony optimization to solve the NP hard multidimensional assignment problem for tracking multiple ground targets.

Reduced palm intensity for track extraction

Reduced Palm intensity function is introduced for track extraction algorithms of filters based on finite point process models to compensate for the spatial, or pair, correlation among detected targets in the Bayes posterior process. Pair correlation function along with intensity and the reduced Palm intensity functions of the Bayes posterior process for the PHD filter are derived. An example is given for which reduced Palm intensity function is used to resolve targets not separated in the posterior intensity function.

Ant colony optimization heuristic for the multidimensional assignment problem in target tracking

Associating measurements with targets is an important step in target tracking. With the increasing computational power, it became possible to use more complex association logic in tracking algorithms. Although itpsilas optimal solution can be proved to be an NP hard problem, the multidimensional assignment enjoyed a renewed interest mostly due to Lagrangian relaxation approaches to its solution. Recently, it has been reported that randomized heuristic approaches surpassed the performance of Lagrangian relaxation algorithm especially in dense problems. In this paper, inspired by the success of randomized heuristic method, we investigate a different stochastic approach, the biologically inspired ant colony optimization to solve the NP hard multidimensional assignment problem.

Application of Topographic State Constraints in Ground Target Tracking

In this work, tracking performance affects of the application of topographic state constraint information was investigated in a ground target tracking scenario utilizing a GMTI radar by using unscented Kalman filter based VS-IMM algorithm as well as VS-SIR particle filter and the tracking results of the aforementioned filters were compared to the results achieved by an unconstrained IMM filter. It was observed that application of topographic state constraints helped the performance of tracking algorithms for tracking onroad targets. However, it was also seen that constrained trackers showed error spikes greater than those of the unconstrained filter when either targets left roads or approached junctions.

Track initiation using multiple bistatic range and range rate measurements with multidimensional assignment algorithm

Track initiation using multiple bistatic range and range rate measurements with multidimensional assignment algorithm was investigated in this work. The assignment algorithm was tested on a problem involving three targets moving in close proximity inside a two dimensional radar network of six bistatic radar pairs. The accuracy and consistency of initiated tracks as well as measurement association performance of the assignment algorithm was demonstrated through Monte Carlo simulations.

Swarm Optimization Approaches for the Generalized Assignment Problem

Colony optimization algorithms have been tested on the generalized assignment problem and their performances have been compared based upon the performance of the auction algorithm in this work. It was observed that both particle colony optimization as well as ant colony optimization methods performed poorly compared to the auction algorithm. Among the heuristics, the PSO algorithm using 1-opt local search has been found to perform better than the other modifications.