Artificial intelligence-enhanced signal detection technique for beyond fifth generation networks

The incorporation of artificial intelligence (AI) has transformed multiple input multiple output (MIMO) technology into a promising candidate for next-generation technologies such as sixth-generation (6G) networks. However, with the increasing number of antennas, it faces significant challenges, and signal detection is one of the main problem. To address this problem, this paper proposes a novel artificial intelligence based integratable detection (AIDETECT) approach for MIMO signal detection. The proposed AIDETECT scheme is an optimized data and model-driven deep learning (DL) approach for MIMO signal detection. The proposed AIDETECT network model is trained, tested, and optimized under different MIMO and massive MIMO (ma-MIMO) use-case scenarios to obtain optimal simulation results. Additionally, this paper presents an in-depth comparative analysis of the optimal selection of hyperparameters for enhancing the optimized performance of the proposed AIDETECT model. The proposed AIDETECT model’s performance is analyzed by benchmarking it against other state-of-the-art AI-based and conventional MIMO detectors. The simulation results show that the proposed AIDETECT schemes perform better than the other AI-based and conventional MIMO detectors. The proposed AIDETECT scheme achieves 63.5% to 87.88% better performance for MIMO and 53.12% to 84.44% for ma-MIMO scenario against AI-based MIMO detectors and also achieves from 88% to 99.9614% better performance than the conventional MIMO detectors for initial simulating MIMO scenarios. Additionally, for computational complexity analysis, simulation results show that the AIDETECT scheme has much lower computational complexity compared to the other AI-based MIMO detectors and conventional MIMO detectors. • Proposed novel AI signal detection for optimizing MIMO and Massive MIMO system. • Paper reviews conventional MIMO detectors, AI methods and emphasizes limitations. • A data production algorithm was proposed for dataset creation. • This study evaluates complexity in MIMO detectors based on complexity. • Proposed method optimizes the hyperparameters for different use case scenario.