SQHCP: Secure-aware and QoS-guaranteed heterogeneous controller placement for software-defined networking

As the large-scale application of software-defined networking (SDN), one major research challenge for the SDN is to place multi-controllers reasonably. The researchers have proposed a lot of good solutions to the controller placement problem (CPP). However, they usually chose the same type of controllers (called homogeneous controllers) and ignored the common-mode fault caused. To fill this gap, considering the heterogeneity of controllers and QoS, this paper first focuses on different types of controllers (heterogeneous controllers) and proposes a Heterogeneous Controller Placement Problem (HeCPP) to minimize network delay, balance heterogeneous controller utilization, and lower control plane fault rate. We introduce a Secure-aware and QoSguaranteed Heterogeneous Controller Placement (SQHCP) approach to solve HeCPP effectively. Specifically, SQHCP consists of two steps. In step 1, the types and the number of heterogeneous controllers are determined based on dynamic planning to enhance the security of the control plane. After that, with the inspiration of the divide and conquer strategy, in step 2, the network is partitioned into several subnets based on the K-means clustering, and the genetic algorithm is improved to place the heterogeneous controller for each subnet to guarantee QoS. The theoretical analysis has proven the feasibility of SQHCP. Simulation shows that SQHCP not only degrades the control plane fault rate but also outperforms the existing approaches in terms of delays and load balancing.

Automated summary

This paper addresses the issue of heterogeneous controller placement problem in software-defined networking (SDN), proposing a Secure-aware and QoS-guaranteed approach to minimize network delay, balance controller utilization, and reduce control plane fault rate. The method involves determining types and numbers of heterogeneous controllers dynamically for enhanced security, partitioning the network into subnets using K-means clustering, and improving genetic algorithms for controller placement to ensure Quality of Service (QoS). Simulation results demonstrate the effectiveness of the approach in reducing control plane faults, improving delay, and enhancing load balancing compared to existing methods.