Optimal Restricted Boltzmann Machine based Traffic Analysis on 5G Networks

There is an increasing demand for accurate traffic prediction techniques to support quality of service optimization, effective network management, and resource allocation with the emergence of Fifth Generation (5G) networks. Artificial intelligence (AI) can able to address the difficulties and complexities of 5G networks. In order to meet future network requirements, it is essential to forecast network metrics in this paradigm. Vehicle-to-everything (V2X) networks are capable of wireless transmission approaches but traffic data interchange in an intelligent transportation system (ITS) continues to be problematic, for example, V2X communication congestion if the vehicles occur unexpectedly in region. This study develops a new cuckoo search optimization with Restricted Boltzmann Machine based Traffic Analysis (CSORBM-TA) for 5G-assisted V2X networks. The goal of the CSORBM-TA technique lies in the optimal prediction of traffic in the V2X network using DL model. To attain this, the CSORBM-TA technique involves three phases. At the preliminary stage, the CSORBM-TA technique applies normalization process. Secondly, the RBM model for the prediction of traffic flow. Finally, the CSO algorithm was used for the parameter tuning of the RBM model. A comprehensive result demonstrates superior prediction accuracy compared to traditional prediction models. The CSORBM-TA model effectively captures and predicts complex traffic patterns, enabling network operators to anticipate future network conditions and make informed decisions for resource allocation and network optimization. Therefore, the CSORBM-TA technique facilitates proactive decision-making, allowing for proactive congestion management, load balancing, and adaptive resource allocation.