SINR-Driven Cross-Layer Energy Optimization (SD-CLEO) for 5G-Enabled Vehicular Sensor Networks

Abstract

Vehicular Sensor Networks (VSNs) represent a critical pillar of the Internet of Vehicles (IoV), enabling continuous Vehicle-to-Everything (V2X) communication for Intelligent Transportation Systems (ITS), autonomous driving, and road safety. However, high mobility and dense vehicle deployments in 5G-enabled environments cause frequent channel access contention, severe co-channel interference, and rapid battery depletion of On-Board Units (OBUs). Centralized Signal-to-Interference-plus-Noise Ratio (SINR)-based power control methods present single points of failure, excessive signaling overhead, and poor adaptability to rapidly fluctuating topologies. To address these challenges, this paper presents the SINR-Driven Cross-Layer Energy Optimization (SD-CLEO) framework  a fully decentralized, three-zone cross-layer architecture operating across the Medium Access Control (MAC) and Physical (PHY) layers: (1) an Adaptive MAC Back-off Scheme (Zone 1) that dynamically adjusts the contention window based on real-time SINR thresholds; (2) a Closed-Loop Transmission Power Control (Zone 2) mechanism that dynamically scales RF output power based on path loss, proximity, and receiver target SINR; and (3) a Proactive Mid-Flight Frame Abort Mechanism (Zone 3) that senses link quality during active packet delivery and truncates corrupted, unrecoverable frames at 0.2 ms rather than the full 1.0 ms duration. Multi-round Monte Carlo simulations demonstrate that SD-CLEO achieves up to an 11.5-fold (1,050%) increase in network energy efficiency compared to standard IEEE 802.11p CSMA/CA, alongside, at 10 vehicles/km, a 30.1% improvement in Packet Delivery Ratio (PDR) and a 13.8% reduction in collision frequency, while keeping the average MAC-level end-to-end latency at about 1.0 ms (at most 1.016 ms), well within the 10 ms end-to-end latency requirement for cooperative collision avoidance in 3GPP TS 22.186.