Detection of intruder using the improved dragonfly optimization algorithm

Abstract

Now a days, we are completely dependent on the systems and also on the internet. Most of our valuable information is stored in the system or in the cloud. So, it’s very necessary to protect our data and also our system from the outside intruders and also from the intruders already entered into our system. To protect our system, intrusion detection systems are constructed using many traditional techniques. Some of them are decision tree, KDD, data mining techniques, Artificial Neural Network. But there are many flaws in these systems. In this paper, we have used the improvised dragonfly optimization algorithm. Here the traditional dragonfly optimization algorithm is improved by adding the convergence and fitness function. This usage of improvised dragonfly optimization algorithm in detecting the intruders has given the best performance than the other methods. To justify the statement, we have shown the experimental study and also comparative study with other two optimization algorithm. Here first the detection rate is computed on the darknet 2020 dataset by using the traditional dragonfly algorithm, then the detection rate is again computed on the same dataset but with the improved the dragonfly optimization algorithm. Finally, in the comparative study it’s very clear that the improved dragonfly optimization algorithm produced the accurate result than the other optimization algorithm.