Turbo-coded type-based detection in man-made noise

This paper considers the type-based detection and turbo decoding in direct-sequence spread-spectrum (DS-SS) systems with totally uncertain man-made noise. We compare the type-based receiver with decoding to several commonly used detectors in AWGN and Middleton class A noise. A decision feedback receiver is proposed to cope with limited training sequence length. Using class A noise model, we show by simulations that the type-based receiver is robust to the number of interferers and their distances from the victim receiver. Close to optimal performance of the type-based receiver is obtained regardless of ambiental noise, number of interferers and their distances from the victim receiver.