دانلود رایگان مقاله انگلیسی الگوریتم بهینه سازی کلونی مورچه مبنی بر اتوماسیون سلولی برای مقابله با حملات DDoS در VANETs - الزویر 2018

ABSTRACT

In Vehicular Ad hoc NETworks (VANETs), reliable data dessimination between vehicular nodes necessitate maximum degree of colloboration as they play a significant role for ensuring the core objective of communication. But the malicious action of vehicular nodes may distrupt the established degree of co-operation as they lead to poor performance inspite of high resource utilization. The malicous activity of vehicular nodes like DDoS attack must be detected and resolved in a potential way by identifying optimal nodes and optimal paths using cellular automata that modifies the search ability in a global way for increasing the speed of convergence. To prevent the search from falling into local optimum point and to enhance the convergence speed, global searching potential, an enhanced version of ant colony optimization (ACO) algorithm called Cellular Automatabased Improved Ant Colony-based Optimization Algorithm (CA-IACOA) is propounded. In CA-IACOA, the update rules and dynamic adaptive adjustment technique in pheromones of the traditional ACO is enhanced to a significant level based three threshold level of tolerance.This enhancement in the adopted pheromoneis mainly for attaining better quality solution through reliable and dynamic increment of pheromone that considers the past history related to the visited paths of the ant agents. Further significant tradeoff that exists between solving quality and solving efficiency is resolved using updated dynamic evaporation factor strategy that aids in rapid convergence. Furthermore, trusted boundary symmetric mutation strategy is incorporated in CA-IACOA for improving mutation and to strengthen mutation efficiency. The simulation experiments of CA-IACOA infers that they are potential in handling DDoS attacks as they obtain quality solutions in identifying optimal nodes and optimal paths for reliable routing.

Conclusion

CA-IACOA is propounded and analyzed for isolating the concept of stagnation which is the major drawback for CA-ACOA mitigation algorithm for facilitating a global searching environment based on Pheromone adaptive adjustment strategy, dynamic evaporation factor strategy and Boundary symmetric mutation scheme through dynamic update rule for reliable packet dissemination between vehicular nodes.CA-IACOA incorporates the advantages of Pheromone adaptive adjustment strategy by replacing the pheromone intensity constant which is a real dynamic changing variable function.CA-IACOA achieves this dynamic adaptation through negative feedback concept for identifying optimal solution. It confirms a trusted and accurate global search dimension for enhancing the degree and probable extent of attacker’s mitigation. It is also proved that the accuracy in mitigation facilitated by CA-IACOA considerably increases as the pheromone intensity constant is increased systematically. The performance analysis of CA-IACOA with CA-IACOA, CA-ACOA, CA-GA and CA-PSO confirms that it is potential in minimizing the prediction variance and delay to a maximum level of 17% and 21% under the influence of increasing numbers of vehicular nodes. It is concluded that CA-IACOA is optimal in its searching performance with respect to multi-modal function Schwefel-2.26 as it facilitates the global optimal point of convergence at an average rate of 19% than the compared Quartic, Exponential and Sumsquare multimodal functions. In the near future, it is planned to devise a mitigation algorithm based on Artificial Bee Colony algorithm using Exponential and Erlang operator distribution for improving the detection and mitigation rate.