Dual indexing-based Spatial Hierarchical Modulation for beyond 5G wireless systems using Deep Neural Networks

Abstract

Increasing spectral efficiency while reducing RF design complexity remains a significant challenge in current 5G MIMO wireless transmission techniques. For this, Spatial modulation has been a potential solution for RF complexity reduction. This work proposes Dual Indexing-based Spatial Hierarchical Modulation (DI-SHM), which incorporates a constellation combiner scheme using hierarchical modulation (HM) with three priority levels to transmit data via spatial modulation with dual antenna indices and variable data mapping. The proposed modulation technique is demodulated using SHM–Deep Neural Network based decoder at a good recovery rate, and the results are compared with the existing techniques. In comparison with the existing literature, result analysis shows that at an SNR of 10 dB, DI-SHM achieves a BER close to (Formula presented.) due to its layered hierarchical modulation structure and dual-antenna indexing. The Prioritization of HM bits by varying the priority levels have also been analysed. This novel technique doubles the spectral efficiency by 12 bits per channel use with reduced receiver complexity.