Real-Time Self-Adaption of Network Security Mechanisms for Dependable Distributed Systems

Use of complex encryption parameters to attain strong encryption results enhances data integrity during transfer between nodes in distributed environments. However, the benefit comes at a performance cost as it increases resource usage. Use of network-based intrusion detection systems allows intelligent interventions where complex encryption parameters are reduced. We propose a method for implementing adaptive adjustment of encryption parameters in the AES algorithm on a client and server platform. Precisely, an algorithm used to implement adaptive encryption is proposed for distributed systems, which optimizes both performance and dependability. In addition, we present an efficient decryption protocol to reverse adaptive encryption. Results show that the overall performance of the proposed algorithm can be monitored, keeping such performances at acceptable levels whilst security of data in transit is still maintained. System performance can also be increased when communication time differences between nodes in a distributed environment are captured as parameters to trigger adaptation of the algorithm to either use more or lesser number of encryption rounds.