Publication

Abstract : User activity logs can be a valuablesource of information in cloud forensicinvestigations; hence, ensuring the reliabilityand integrity of such logs is crucial. Mostexisting solutions for secure logging aredesigned for conventional systems rather thanthe complexity of a cloud environment. In thispaper, we propose the Cloud Log AssuringSoundness and Secrecy (CLASS) process as analternative scheme for the securing of logs in acloud environment. In CLASS, logs areencrypted using the individual user’s public keyso that only the user is able to decrypt thecontent. In order to prevent unauthorizedmodification of the log, we generate proof ofpast log (PPL) using Rabin’s fingerprint andBloom filter. Such an approach reducesverification time significantly. Findings from ourexperiments deploying CLASS in OpenStackdemonstrate the utility of CLASS in a realworldcontext.