An Intersection-Based QoS Routing for Vehicular Ad Hoc Networks With Reinforcement Learning
Lanlan Rui,Zhibo Yan,Zuoyan Tan,Zhipeng Gao,Yang Yang,Xingyu Chen,Huiyong Liu
DOI: https://doi.org/10.1109/tits.2023.3271456
IF: 8.5
2023-01-01
IEEE Transactions on Intelligent Transportation Systems
Abstract:Vehicular ad hoc networks (VANETs) have the characteristics of high mobility, frequently changing topology and uneven distribution, which made it a challenge to design an efficient and robust routing protocol with low latency and high packet delivery rate. Currently, intersection-based routing method and full-path based routing method are two popular solutions for the packet routing in VANETs. Although the intersection-based routing method has better real-time performance, it has the problem of local optimization, making the routing results not global optimal. Although the full-path based method can obtain the global optimal solution, it is weak in dealing with the dynamically changing topological network. Aiming at solving the above problems, this paper designs an intersection-based QoS routing (IQRRL) algorithm, which mainly includes two crutial steps: the next intersection selection and the next hop vehicle selection. In the selection of the next intersection, this paper uses an improved intersection-based routing protocol. In addition to considering connectivity and delay, IQRRL also considers the communication quality from the neighbor’s road to the destination node while evaluating the quality of the neighbor’s road, which minimizes the problem of local optimization. When the next intersection is determined, a road is then determined, and then the next hop vehicle within the road should be chosen to relay the packet forward. In the next-hop vehicle selection step, this paper adopts multi-hop evaluation technology based on reinforcement learning. In addition to using “greedy decision-making” to select the next-hop vehicle, it also comprehensively evaluates whether the next-hop vehicle is still optimal in the future, so that the stability and reliability of data forwarding are improved and local optimal problems are avoided. Besides, this article uses a simulation system to compare IQRRL with other routing algorithms. The result reveals that IQRRL outperforms in terms of packet delivery ratio and transmission delay.
engineering, electrical & electronic,transportation science & technology, civil