Certain investigations on file storage security in cloud to ensure confidentiality and availability

Abstract

Data is a valuable asset and when this data is moved to the cloud it becomes a great concern that needs to be preserved and protected. The data stored in cloud are of different types and each type of data requires diverse degree of security and protection. Data basically requires confidentiality, integrity and availability to be maintained at all stages. Data confidentiality is all about data privacy where only authorized users can gain access to the data. Data integrity is about the consistency and accuracy of data such that the users receive unaltered data content anywhere, everywhere. Data availability deals with protecting the data from unwanted malicious events that is natural or intentional and making the data service available to users irrespective of any situation. This research primarily focuses on investigating data confidentiality and availability issues, it s possible threats and existing solutions. This basic idea has paved the way for enhancing data confidentiality by developing Intelligent Division Model (IDM) and Organized Node Selection (ONS) mechanism where data file is taken as input and is dispersed into fragments by splitting the file based on the fragmentation threshold generated through this IDM. The similarity between data fragments are examined via WordNet ontology that supplies the meaning of data in each fragment and the fragments are finally formed in such a way that compromising one data node will not reveal the content of the entire file. For recovery purpose, a fact file is generated that holds the number of bits and the position of each bit in the file. It is XORed with a random number and is stored on dedicated servers. newline