A delay-constrained node-disjoint multipath routing in software-defined vehicular networks

Data routing processing in vehicular ad hoc networks (VANETs) with some QoS constraints, including delay bound and limited bandwidth, is an NP-Hard problem. Routing efficiency can be enhanced if more information is applied in routing decisions. Software-defined networks (SDNs) optimize network resources such as bandwidth by providing a global perspective of vehicular network topology, improving routing performance. This paper proposes a new method using the SDN integrated into vehicular networks. In this method, the controller nodes compute the r-least-cost node-disjoint path by having a global view of vehicles' information and trade-offs among the delay bound and path utility. The path utility factor is calculated based on the vehicles' residual bandwidth, packet congestion level, and link stability. In addition, according to the application type, packet size, requirement bandwidth, and delay bound of application, a priority-based scheduling algorithm is proposed. The proposed method calculates disjoint paths by considering the path utility factor and prioritizing the applications in data routing processing, which can react appropriately to increased traffic, continuous topological changes, and link interruptions in VANETs. The efficiency of the proposed method is superior against other methods in the literature in scenarios with different vehicle densities and various traffic loads in terms of data packet delivery fraction, average end-to-end delay, throughput, and normalized routing load through simulation.