MAMP Detector for AFDM under Doubly Dispersive Channels

With the increasing demand for multi-carrier communication in high-mobility scenarios, it is urgent to design new multi-carrier communication waveforms that can resist large delay-Doppler spreads. Affine frequency division multiplexing (AFDM) is a promising solution for next-generation wireless access to achieve strong delay-Doppler spreads resilience in possible scenarios. By adjusting AFDM parameters to match the multipath delay and Doppler shift of the doubly dispersive channels, AFDM can obtain optimal diversity gain. However, AFDM encounters symbol dispersion in the underlying modulation domain, i.e., the discrete affine Fourier transform (DAFT) domain, which poses significant challenges for signal detection. To overcome this challenge, we proposed an efficient detector for AFDM based on the memory approximate message passing (MAMP) algorithm. It uses a memory matched filter (memory MF) to achieve a high-precision reconstruction of the original signal at the cost of low complexity. We compare the bit error ratio (BER) performance of the proposed MAMP and traditional linear minimum mean square error (LMMSE) detector, and simulation results show that MAMP significantly outperforms LMMSE under integral delay-Doppler conditions.