PISA, ITALY - The integration of artificial intelligence (AI) and machine learning (ML) with fifth-generation wireless communication technology (5G) enables the anticipation of traffic patterns within a limited range by examining geographical data, engineering variables, and records. 

This analysis facilitates prediction of peak traffic volumes, resource allocation, and identification of specific application categories. AI and ML techniques also forecast traffic patterns and needs, assess channel conditions and quality, enhance quality of service (QoS) performance, and optimize resource use. AI algorithms analyze data from interconnected vehicles and traffic control systems to clear congestion, improve safety, and optimize routes for individual vehicles. The issue of predicting 5G network traffic should thus be reframed as a challenge of time-series prediction modeling. Previous studies mostly targeted two distinct approaches. By harnessing the abilities of AI and ML in conjunction with 5G, operators are switching from network management to service management in a shift upending network operations and maintenance.

ML, AI techniques apply to beamforming

Implemented through millimeter-wave technology, 5G uses beam-based cellular coverage, unlike 4G, which relies on sector-based coverage. ML enables a 5G cell site to calculate a group of potential beams - from either the serving cell site or its nearby cell one. An optimal set minimizes the number of beams, while maximizing the likelihood of capturing the best one. The optimal beam shows the highest signal strength - also referred to as Reference Signal Received Power. The likelihoo