SDN-Assisted Low-Latency Resource Management for V2X Communication in Heterogeneous Cellular Networks

Abstract

Vehicle-to-Everything (V2X) communication requires low-latency, reliable, and context-aware connectivity across vehicle-to-vehicle, vehicle-to-infrastructure, vehicle-to-pedestrian, and vehicle-to-network services. Heterogeneous cellular networks can support these services by combining 5G New Radio, cellular sidelink, roadside units, Wi-Fi or DSRC links, and multi-access edge computing; however, high mobility, variable traffic density, frequent handovers, congestion, and static resource-allocation policies can degrade end-to-end delay and packet delivery. This paper presents a software-defined networking assisted framework for latency-aware resource management in heterogeneous V2X environments. The proposed framework separates network control from forwarding, maintains a global view of vehicular network state, classifies V2X flows by service criticality, and dynamically selects routes and bandwidth allocations using delay, congestion, handover, and priority constraints. A mathematical formulation is developed for minimizing weighted end-to-end latency under link-capacity, minimum-bandwidth, packet-delivery, congestion, and controller-processing constraints. The paper also defines the Latency-Aware SDN Resource Allocation Algorithm for V2X Networks and specifies a reproducible simulation methodology using NS-3/SUMO or OMNeT++/Veins with multiple baselines, confidence intervals, and ablation analysis. Because the source report did not include raw simulation data, the revised results section includes clearly labelled hypothetical numerical plots and reporting values only to demonstrate result formatting and interpretation; these values are not validated empirical findings. The study provides a rigorous foundation for validating SDN-MEC orchestration in next-generation vehicular communication systems