A Non-Orthogonal ISAC Architecture based on OCDM-IM and Superimposed Training with Iterative SIC Receiver

This paper proposes a non-orthogonal approach for integrated sensing and communication by superimposing a Zadoff-Chu (ZC) sequence onto an orthogonal chirp division multiplexing with index modulation signal. The ZC sequence acts simultaneously as a radar signal and a pilot for communication channel estimation (CHEST). To mitigate mutual interference, we design an iterative receiver that uses soft-decision successive interference cancellation and a linear minimum mean square error (MMSE) estimator. Numerical results show that this significantly reduces the CHEST MSE while efficiently maintaining data rates. The system achieves a peak-to-sidelobe ratio of 24.3 dB at a 20 dB signal-to-noise ratio, with centimeter-level range accuracy, confirming its effectiveness for future sixth-generation applications.