Sparse Channel Estimation and Data Detection for AFDM With Superimposed Pilot Scheme

Affine frequency division multiplexing (AFDM) is a multi-chirp waveform proven to be effective for high-mobility communications in next-generation wireless systems. However, the existing embedded pilot-aided (EPA) scheme performs transmission surrounded by null guard samples on each side and the spectral efficiency (SE) is limited. To address this issue, we introduce a novel transmission framework, the superimposed pilot (SP) scheme, which aims to enhance the robustness of AFDM systems against time-varying channels and improve SE by fully integrating data symbols with pilots. By leveraging the sparse nature of time-varying channels, we reformulate the channel estimation problem as a sparse signal recovery task. An orthogonal matching pursuit (OMP) algorithm is considered for channel estimation within this framework which exploits AFDM channel sparsity, and consequently has low computational complexity. Additionally, we propose a strategy by iterating between channel estimation and data detection. Simulation results show that the bit error rate (BER) performance of the SP framework approaches that of perfect channel estimation and significantly surpasses the EPA scheme with the same SE.