A Novel Simulated Annealing based Routing Algorithm in F-SDNs

In recent years, routing algorithms have attracted tremendous research attentions in Flying Ad hoc Networks (FANETs) in order to provide efficient networking to fulfill the demand of Intelligent Transport System (ITS). Due to the lack of a global network information in FANETs, existing routing algorithms still have many issues in finding the optimal next-hop node, suffering low packet delivery rate and high delay. With the inclusion of Flying-Software Defined Network (F-SDN), providing global network information has eased the routing and forwarding of data for Unmanned Aerial Vehicles (UAVs). In this paper, we introduce a new routing algorithm by our Modified Simulated Annealing (MSA) algorithm to increase performance of networking in dense environments. The experimental results show the proposed algorithm outperforms the standard algorithm in the finding of routing paths in F-SDNs.

I. Introduction

The recent developments of UAV systems introduce high demands of consumers for their ease and safe communication brought the necessity for improvements in Intelligent Transportation Systems (ITS) on managing and organizing UAV tra□c. As a primary sector of such a network, Flying Ad hoc Network (FANET) enables communication with the Drone to Drone (D2D) communication, as well as Drone to Infrastructure (D2I) communication. The dynamic nature of FANETs leads to the change of network topology and network density. In this context, routing in FANET plays an important role in information propagation which should overcome the drawbacks brought by such a dynamic environment. Even though the existing routing protocols are designed according to the above characteristics, many routing protocols still fall short in addressing issues with regards to finding the optimal routing path. The assurance for transferring data to a node with better link stability and shorter distance to the destination is very important. Part reason of this is the lack of a global network information in FANETs. Consequently, failing in finding an optimal route of data messages could result in the existing to be unstable routes. This makes a vulnerability in packet delivery which can further result in packet drops [1]. To further utilize the routing algorithms in FANET, Software-Defined Network (SDN) paradigms can be integrated with FANETs to form Flying-Software Defined Networks (F-SDNs) which can make it a very versatile networking platform. Softwarized networking enables the decoupling of control and data planes of F-SDN, which can further lead to (1) an abstraction for applications to the underlying networking infrastructure, and (2) a logically centralized networking intelligence and states [2]. The advance demands of FANETs include high data throughput, scalability, latency and high mobility in which these demands could be met by SDN easily due to the advancements of SDN. It has been shown that softwarized networks can provide a better overview of the network topology enabling a more centralized computation for tra□c management for multiple routing algorithms [3]. Also, a programmable controller can simplify the network architecture and lay out the foundations for highly flexible networking infrastructure [4].