A Context-Aware Predictive Access Control Scheme for Massive IoT Devices in Smart City Environments

The rapid digitization of urban environments has led to a significant increase in the deployment of IoT devices, resulting in severe access congestion when many devices want to access the network simultaneously. To address this challenge, we propose the Context-Aware Predictive Access Class Barring (CAP-ACB) scheme. This scheme updates the ACB factor, which is critical for the random access process, in each time slot. It does so based on three key factors: prediction, feedback, and context awareness. Essentially, in every time slot, the ACB factor is improved by considering these parameters, which include the estimation of new device arrivals, feedback from idle preambles, and insights from rush hour patterns. Additionally, we consider the Msg3 splitting method to tackle collision issues. Simulation results demonstrate that CAP-ACB achieves higher throughput and lower latency, compared to fixed ACB and other traditional random access methods. These findings highlight CAP-ACB as an effective and practical solution for managing massive IoT connectivity in smart city environments.