Supervised Learning Resource Allocation Over Bandwidth Constrained Linear Vehicular Network

Vehicular networks demand efficient resource allocation due to dynamic traffic-related demands. Fixed allocation risks service quality and wastage. The paper focuses on two random forest models for dynamic access point resource allocation which were trained using dataset generated by priority algorithm. Model I, allocating resources individually to nodes, showed disparities between predicted and expected allocations (R-squared: 0.808). In contrast, Model II, aggregating demands based on priority levels, demonstrated superior accuracy (R-squared: 0.989), highlighting its effectiveness in resource allocation.