Evolving Unimodal Secure Load Balancing In Heterogeneous Clustered Web Servers with Nature Inspired Meta Heuristic Scheduling Scheme

Abstract

At present, there is a massive usage of internet sources with the implication of distributed system architectures. Information exchange with several web applications requires security methods. In particular, sensitive data faces the risk of security threats. With an intruder act in the network causes security breaches despite using various security providing algorithms. Though there exist security algorithms like RSA, AES, DES, etc., many vulnerabilities occur due to static/dynamic key usage. Attacks like brute force, collusion, SQL Injection are some of those which keep data in a vulnerable state. Key generation is the primary concept in these algorithms, and when the key can be achieved through trials, it is much more vulnerable. The performance of the security algorithm depends on the key size. Computation and time complexity increases when the key size is large. Cryptography combined with a biometric feature is an enhanced security mechanism developed in this research work. Extracting Ear feature, encoding with RNA, and generating a dynamic key with Ntru cryptosystem resolves the security issues. A random 512-bit secret key is generated from the ear feature in the developed methodology. As the methodology internally encodes the generated key with RNA sequencing, it is highly impossible to guess through various key cracking approaches. User satisfaction and requirement lies in measuring server availability. When much loads of data are in a communication network causes the overhead of adaptability, accessibility, and imbalance in load distribution. Traditional load balancing in clustered web servers are limited in faster response times, and greater performance. Overloading in the clusters can be handled by an efficient load balancing mechanism. newline