Detection and mitigation of cloud based distributed denial of service attacks

Abstract

Cloud computing provides a convenient technique to obtain services, resources and applications across the Internet. Research shows that different kind of DDoS attacks on cloud result in different effects. This thesis proposes a novel architecture that combines a well posed stacked sparse AutoEncoder (AE) for feature learning with a Deep Neural Network (DNN) for classification of network traffic into benign traffic and DDoS attack traffic. AE and DNN are optimized for detection of DDoS attacks by tuning the parameters using appropriately designed techniques. The improvements suggested in this thesis lead to low reconstruction error, prevent exploding and vanishing gradients, and lead to smaller network which avoids overfitting. A comparative analysis of the proposed approach with ten state-of-the-art approaches using performance metrics-detection accuracy, precision, recall and F1- Score, has been conducted. Experiments have been performed on CICIDS2017 and NSL-KDD standard datasets for validation. Proposed approach outperforms existing approaches over the NSLKDD dataset and yields competitive results over the CICIDS2017 dataset. After the detection of attacks, it is crucial to mitigate these attacks. So, this thesis also proposes a novel filtering-based approach for mitigation of DDoS attacks. This approach provides counters against DDoS attacks launched via IoT based botnets and zombies. The CAPTCHA based on gesture verification is used to filter out bots from humans. The proposed approach will track the behavior of clients accessing the network. The clients gain trust by accepting and replying correctly to the challenge. Only the trusted clients are given access to the network, otherwise they are blocked. The proposed gesture based CAPTCHA approach is more reliable than the other techniques like text based, image based, audio based, puzzle based etc. newline