SECURING GOVERNMENT NETWORKS: ADVANCED TRAFFIC ROUTING AND ANOMALY DETECTION VIA SEAGULL OPTIMIZATION AND DEEP LEARNING WITH SDN

Abstract

The transformation of local self-governance into digital form requires continuous availability of services through the internet, including permit applications, taxation, and access to public documents. Nonetheless, the increase in traffic along with growing threats of DDoS attacks jeopardizes the work of the administrative system where server response time is an important parameter which is often ignored by current load balancers. The research provides the idea of using hybrid Artificial Intelligence for securing the local government networks and optimizing their functioning. More precisely, a combination of Artificial Fish Swarm algorithm along with a CNN-RNN model was used for effective detection of DDoS attacks in software-defined networking. At the same time, a metaheuristic method of seagull optimization is used to calculate fitness function for finding the optimal routes for load balancing in case of high traffic. The empirical evaluation revealed the considerable improvements in terms of Root Mean Square Error: the proposed model achieved a better result compared to CNN (16.13%) and RNN (22.91%). In addition, the new approach increased the network throughput by 31.98% and 33.96% compared to OpenFlow and Linux ProGFE, correspondingly.