Localization of illegal radios utilizing cross-correlation of channel impulse response with interpolation in urban scenarios

Fingerprint-based localization is expected to be superior compared to other conventional range-based localization techniques especially in dense urban scenarios which are generally non line-of-sight (NLOS). However, when we consider localizing illegal radios whose parameters such as bandwidth and center frequency are unknown, there is a high possibility that the initial parameters used to create the fingerprint database in the training phase do not match the illegal radio´s parameters. In this paper, a novel fingerprint-based localization technique which utilizes interpolated channel impulse response (CIR) cross-correlations as fingerprints is proposed, which does not require a priori information of the illegal radio. Results show that the proposed technique is very promising in localizing illegal radios in urban NLOS environments.