A Family of MMSE Detectors for MIMO-OCDM Systems with Near-Optimal Performance

Orthogonal chirp division multiplexing (OCDM) is a promising technique for sixth-generation (6G), offering better resilience against multipath fading and interference compared to orthogonal frequency division multiplexing (OFDM). In this paper, widely linear minimum mean square error (WL-MMSE) detectors are introduced in multiple-input multiple-output orthogonal chirp division multiplexing (MIMO-OCDM) systems to deal with improper complex signals by minimizing noise and interference which outperform traditional MMSE detectors. Building on this, likelihood ascent search (LAS) detectors, known for their near-optimal performance with low complexity, are introduced. LAS detectors use iterative neighborhood search, initialized with MMSE and WL-MMSE outputs, to enhance detection performance by increasing likelihood and reducing error probability. The proposed detectors demonstrate over 5 dB SNR gain compared to MMSE in MIMO-OCDM systems and a 12 dB gain over MIMO-OFDM systems.