Residual self-interference cancellation in NOMA-OFDM full duplex massive MIMO

The presence of self interference (SI) in full duplex (FD) created by the Base Station (BS) severely hampers the performance of communication system, in addition to multiple access interference (MAI). To tackle such impairments, a new scheme based on NOMA-OFDM in Massive MIMO is proposed. Moreover, blind channel estimator is used to provide channel state estimation. This architecture is based on the blind source separation approach (SOBI) algorithm which estimates the residual (RSI) channel in the OFDM-NOMA-FD-Massive MIMO architecture, to combat the RSI and MAI created by the base station and the multi-user presence, respectively. Furthermore, The so-mentioned architecture incorporates ordered successive interference canceler minimum means square Error (OSIC-MMSE) as receiver, in order to offer an additional reliability of the wireless link. A comparative performance study between the proposed architecture, the least square estimator and half duplex (HD) architectures is performed. To the best of author knowledge, the performance of the aforementioned architecture was not previously investigated, more particularly SI power is characterized under Rician distribution. From simulation results, the adopted architecture returns acceptable bit error rate performance.