Modulated Lapped Transform-Based OFDM-IM With Low Complex Deep Neural Network Detector
Ahmed Waheed, Smaya Moher · IEEE Open Journal of the Communications Society · 2026
This paper presents a novel modulated lapped transform-based orthogonal frequency division multiplexing with index modulation (MLT-OFDM-IM) scheme combined with a low-complexity deep neural network (DNN) detector. The main objective is to meet the stringent demands of future industrial wireless networks, including ultra-reliability and low-latency communication. The proposed MLT-OFDM-IM exploits the time-frequency localization properties of the modulated lapped transform (MLT) to show robustness against carrier frequency offset (CFO) and inter-carrier interference (ICI). A closed-form mathematical expression based on pair-wise error probability is derived to estimate the bit error probability (BEP) of MLT-OFDM-IM for performance analysis. Moreover, a low-complexity DNN-based detector for MLT-based OFDM-IM is proposed that can achieves superior performance compared to traditional maximum likelihood (ML) detection under perfect channel state information (CSI) and imperfect CSI. Extensive simulation results demonstrate that the proposed scheme significantly outperforms existing competitive benchmark schemes, such as FFT based OFDM-IM and OFDM. Specifically, at SE of 1.25 bps/Hz and BEP of 10−2, our proposed scheme achieves an additional gain of 3 dB.