USING HIDDEN MARKOV MODEL TOWARDS SECURING THE CLOUD: DETECTION OF DDoS SILENT ATTACKS

Abstract

Cloud Computing has presented itself as a promising solution to new entrepreneurs as well as existing organizations for management of their IT needs at various levels. Many cloud service providers have exposed cloud services at cheap prices, which allow users at all levels of society to materialize their ideas and make them available across the globe. While the response has been overwhelming, the application areas where security of data is of utmost importance have not shown much interest. Hence incorporating dependable security measures in the cloud computing technology would be a good move since the aspect of security has turn out to be one of the main things to consider. In this thesis work we took an initiative as we adopted and/or interpolating Hidden Markov Model into the circles of Cloud Computing, as we exploited it’s capabilities in detecting the silent attack traces that dodge and/or bypass a set of methodical mechanisms intended to sense and prevent them into the cloud system. We modelled our hackers’ attack scenarios where we defined some states and observations. The hackers’ clusters have been taken to be the states, along with the observations which have been taken to a series of Virtual Machines on to which the attacker trades upon from the silent invasion through to the vulnerable and finally the target Virtual Machine. The modelling is given in such a way that, if we have a generated sequence of observed attacks, we must be able to find out the roots of done attackers from the grouped hackers.