On angle estimation in GSM passive coherent location systems

The paper discusses the angle of arrival (AOA) estimation task in GSM based passive coherent location (PCL) systems that bases on the scan pattern maximization method. Being optimal at maximum likelihood (ML) sense for the continuous case, this method may degrade if only discrete values of the pattern are available and the scan rate is low. Based on the evaluation of experimental data this contribution will investigate the enhancement that can be reached if the scan pattern is sampled densely. Two methods are proposed to deal with this task either at raw data level, i.e. through angular scan rate enhancement or at detection level by interpolation. The computational load of both methods is discussed and an alternative processing concept is proposed aiming at reducing redundancies and accelerating the whole processing. Especially, the highly parallel structure of the concept will be exploited for an efficient implementation on graphics processing units (GPU).