Dynamic track cluster algorithm based on subtractive clustering

To the problem of track clustering caused by multipath propagation, an adaptive dynamic track cluster (ADC) algorithm was proposed and two indexes to evaluate the performance of track cluster algorithm were defined. The set of sample was changed dynamically in the procedure of cluster. After the cluster centers were acquired, the ADC algorithm determined the sample's class through the nearest neighbor method. Simulation results show that the ADC algorithm can improve the precision of track cluster dramatically.