An Efficient Approach for Data Security and Integrity in Multi Cloud Computing

Abstract

There is considerable increase in the significance given to Cloud Computing (CC) and it has been
getting a rapidly developing attention in the communities of science and industries. It is easy on privileges,
convenient on-demand access to a repository of resources and services that are configurable. The services and
resources observed on the cloud could be reliably implemented with lesser maintenance expense, effort and
better efficiency. Still, the cloud is regarded to be a computing platform with insecurity from the user
perspective owing to the absence of security with regard to confidentiality and integrity. Shielding techniques
for cloud systems has to be developed so as to protect the sensitive information employing cryptographic
methods and in addition keep up the user information with intactness saved on the remote storage by
safeguarding it against dangerous entities or behaviors. In order to deal with this issue, the available system
proposed a multilevel firefly Algorithm Threshold (MFAT) based secret sharing schema. The generation of the
secret key is done by making use of the firefly algorithm. The key that is generated, is utilized for encrypting
the file. Thereafter the user divides the secret key into several shares and then distributes them between all
resource providers for the purpose of decryption. For improving the level of security and integrity the newly
introduced system presented a Multilevel Dragonfly Optimization Threshold schemes (MDOT). In this
technical work, Dragonfly Optimization algorithm is employed for the generation of key. After this, MDOT
does the encryption of the file exploiting the secret key prior to the distribution of the key shares among the
participating resource providers who are assumed to be trustworthy. At last, Adaptive Neuro Fuzzy Inference
System (ANFIS) based classifier is applied for performing the check on the cloud user data integrity. Results
obtained from experimentation proves that the newly introduced MDOT based secret sharing mechanisms
yield high security in comparison with the available state of art techniques