Answers are grounded in this paper and cite their sources.
Analysis of deep learning methods improves signal detection in spatial modulation systems, indicating enhanced performance metrics like spectral efficiency.
Index modulation (IM) has attracted increasing research attention in recent years. Spatial modulation (SM) as a popular IM scheme is effective to increase spectral efficiency using the antenna index to transmit extra information bits. It can also address some issues that occur in multiple-input multiple-output systems, such as inter-channel interference and inter-antenna synchronization. Artificial intelligence, especially deep learning (DL), has made significant inroads in wireless communication. Recently, more researchers have started to apply DL methods to IM-based applications such as signal detection. Many results have proven that DL methods can achieve breakthroughs in metrics like bit error rate (BER) and time complexity compared to conventional signal detection methods. However, the problem of how to design this novel method in practical scenarios is far from fully understood. This article surveys several DL-based signal detection methods for IM and its variants. Moreover, we discuss the performance of different neural network structures, some of which can achieve better performance compared to original neural network. In the implementation, trade-offs between BER and time complexity, as well as neural network’s training time, are discussed. Several simulation results are provided to demonstrate how the DL method in signal detection of SM can lead to improvements in BER and time complexity. Finally, some challenges and open issues that suggest future research directions are discussed.
No takes yet. Share an insight, caveat, or question.
Jin et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: