A single-packet method for adaptive maximum likelihood CCI identification and mitigation

Joint modeling and detection of a narrow-band digitally-modulated signal and co-channel interference (CCI) signals is considered. A PSP-based algorithm that is compatible with array measurements is described and shown to efficiently detect the desired user´s data while tracking the time-varying intersymbol interference (ISI) channels of the intended user and all CCI users. This algorithm yields improvements in CCI-rejection of approximately 18 dB relative to traditional approaches (i.e., ignoring the CCI) with only a single antenna. An algorithm to estimate the number of CCI signals present and their time spans is also developed. This algorithm, based on ML estimation and the minimum description length (MDL) principle, is shown to correctly identify the CCI environment with high reliability using the training information available in a single TDMA packet