Intelligent Network Slicing in V2X Networks – A Comprehensive Review

Abood, Mohammed Salah; Wang, Hua; He, Dongxuan; Kang, Ziqi; Kawoya, Agnes · 2023 · Crossref

DOI: 10.37965/jait.2023.0208

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This comprehensive review addresses the challenge of optimizing Vehicle-to-Everything (V2X) networks through intelligent network slicing, motivated by the increasing connectivity demands of autonomous systems and the Internet of Things. While 5G networks offer the necessary high throughput and low latency for V2X applications, existing Long-Term Evolution (LTE) systems and traditional network management techniques struggle to meet dynamic requirements without introducing excessive complexity. The paper argues that integrating Machine Learning (ML) and Deep Learning (DL) with network slicing is essential to automate resource allocation, enhance reliability, and support diverse use cases ranging from safety-critical communications to high-bandwidth infotainment. The authors conduct a systematic review of existing literature, proposing a novel taxonomy to categorize V2X slicing research. This taxonomy is structured around four key dimensions: enablers (such as Software-Defined Networking, Network Function Virtualization, and cloud/fog computing), configurations (distinguishing between inter-slicing for different traffic types and intra-slicing for sub-services), specific network requirements (throughput, latency, and bandwidth), and the ML/DL algorithms employed for control and management. The review analyzes various studies from both end-user perspectives, focusing on safety and security, and industry perspectives, focusing on control mechanisms for enhanced mobile broadband, massive machine-type communications, and ultra-reliable low-latency communications. Key findings highlight the efficacy of ML algorithms in managing the complexity of V2X slices. The review details how reinforcement learning, Q-learning, and deep reinforcement learning are utilized to dynamically allocate resources, minimize interference, and optimize service quality in both inter- and intra-slicing configurations. For instance, hybrid approaches combining SDN with fog computing are shown to improve throughput and reduce latency compared to centralized cloud solutions. The paper also contrasts traditional optimization methods, such as game theory and genetic algorithms, with ML-based approaches, noting that while traditional methods offer interpretability, they often suffer from high computational loads and poor adaptability to dynamic network conditions. The significance of this work lies in providing the first comprehensive survey that applies ML algorithms to V2X slicing from a dual perspective, filling a gap in existing literature that often focuses on isolated slicing areas. The authors identify critical challenges, including limited training datasets, model complexity, security vulnerabilities, and the interpretability of ML models. The paper concludes by outlining a future roadmap for research, emphasizing the need for low-computation techniques, robust dataset collection, and improved control mechanisms to ensure the scalability and security of intelligent V2X networks.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• v2x connected vehicle