Research on the Controller Deployment Strategy of Aviation Information Network

Abstract With the development of intelligent aviation platforms, Aviation Information Networks (AIN) has emerged as a promising architecture to accommodate various aviation missions. However, the current Airborne Ad Hoc Networks (AANETs) suffer from the challenges of tightly-coupled devices and complex aviation environments. To tackle the issues, the Software Defined Network (SDN) is introduced to manage the AINs by decoupling the control plane from the data plane and allocating the network resources logically. Considering the large-scale and dynamic changing characteristics of the AINs, we utilize the hybrid architecture consisting of multiple controllers to establish the control system. To construct the AIN of the SDN paradigm, the first issue that needs to be reckoned with is the controller deployment. This paper proposes a controller deployment strategy that includes network division and regional deployment. Firstly, the network is rapidly divided into areas according to the key performance indicators of the network, and then the platform with the highest reliability is selected as the controller to be deployed in each area. The experimental results show that the proposed algorithm has low computational complexity and can effectively improve the network performance in terms of average requirement delay, load balance index, controller deployment cost, and reliability, which is suitable for addressing the controller deployment issue in aviation scenarios.