CSI-based NOMA with AFDM for ISAC in Doubly Dispersive Channels

With the increasing demand for massive Internet of Things (IoT) connectivity in high-mobility scenarios, concerns arise regarding coexistence, scheduling, and spectrum scarcity due to additional sensing signals. This paper proposes a novel affine frequency division multiplexing (AFDM)-based non-orthogonal multiple access (NOMA)-integrated sensing and communication (ISAC) uplink transmission scheme tailored for doubly dispersive channels. We consider the independent communication user (CU) transmitting pure data with ideal channel knowledge, while the sensing user (SU) employ a sparse multiple pilots embedded (MPE) scheme. To mitigate signal overlapping, an iterative channel estimation (ICE) algorithm leveraging AFDM channel state information (CSI) is developed. This algorithm enhances signal separability, achieves accurate SU channel estimation, and aids in reliable CU data detection. Numerical results demonstrate significant improvements in SU channel estimation mean squared error (MSE), CU communication bit error rate (BER) and spectral efficiency with an increasing number of iterations.