A Mobility-Aware Dynamic TDD Scheduling Framework for 5G NR-Based V2X
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
Efficient uplink and downlink resource allocation is critical for low-latency and high-reliability communication in 5G enabledVehicle-to-Everything (V2X) environments. Traditional Time Division Duplexing (TDD) configurations rely on static or semi-static slotscheduling, which fails to adapt to the rapidly changing mobility and communication demands of vehicular networks. This paper presentsa mobility-aware dynamic Time Division Duplexing (TDD) scheduling framework designed to adaptively allocate uplink and downlink(UL/DL) transmission resources in response to real-time traffic dynamics. The proposed framework integrates a lightweight learning-basedscheduler with a zone-partitioned urban mobility model, leveraging contextual features such as vehicle density and latency feedback todriving slot allocation decisions. Built on the SUMO mobility simulator and TraCI interface, the system emulates a realistic Manhattan-griddeployment and supports fine-grained UL/DL scheduling across dynamically changing zones. Comparative evaluations against load-basedand latency-aware baselines demonstrate significant improvements in uplink latency and delivery reliability. The results confirm that theproposed framework meets stringent URLLC requirements while maintaining scalability and real-time feasibility for next-generation urbanV2X deployments.
Abstract
Efficient uplink and downlink resource allocation is critical for low-latency and high-reliability communication in 5G enabledVehicle-to-Everything (V2X) environments. Traditional Time Division Duplexing (TDD) configurations rely on static or semi-static slotscheduling, which fails to adapt to the rapidly changing mobility and communication demands of vehicular networks. This paper presentsa mobility-aware dynamic Time Division Duplexing (TDD) scheduling framework designed to adaptively allocate uplink and downlink(UL/DL) transmission resources in response to real-time traffic dynamics. The proposed framework integrates a lightweight learning-basedscheduler with a zone-partitioned urban mobility model, leveraging contextual features such as vehicle density and latency feedback todriving slot allocation decisions. Built on the SUMO mobility simulator and TraCI interface, the system emulates a realistic Manhattan-griddeployment and supports fine-grained UL/DL scheduling across dynamically changing zones. Comparative evaluations against load-basedand latency-aware baselines demonstrate significant improvements in uplink latency and delivery reliability. The results confirm that theproposed framework meets stringent URLLC requirements while maintaining scalability and real-time feasibility for next-generation urbanV2X deployments.