New Joint Channel Estimation and Interference Cancellation Method for FBMC Systems Based on Recurrent Neural Networks

Abstract Filter bank multicarrier (FBMC) is an important system used in fifth generation (5G) networks to maximize available bandwidth while meeting high spectral efficiency requirements (SE). Multicarrier modulation (MC) is an alternative modulation method used by FBMC. It is a viable alternative to the orthogonal frequency division multiplexing (OFDM) modulation method. This paper focuses on the joint channel estimation and interference cancellation (JCEIC) in FBMC systems. Recurrent neural networks (RNN) are used to estimate the ideal channel and get back the correct transmitted signal with low BER. We estimate the channel for doubly-selective channels using scattered pilots in the time and frequency correlation, and we use low-complexity interference cancellation. For JCEIC, RNN is proposed. The JCEIC algorithms' output sequences are used as inputs for the RNN. The simulation results show that the suggested technique is close to the ideal channel and has a higher BER than the other earlier methods.