Data centric approach to analyzing security threats in Underwater Sensor Networks

The concept of securing data in Underwater Sensor Networks (UWSN) is an ongoing conflict due to the many challenges faced in this harsh environment. Due to the restriction of energy among underwater modems, we must ensure that our transmissions are efficient as possible. Furthermore, we must consider the open possibility of mobile malicious nodes interjecting false packets into our network. Information-centric architectures proves to be a potential solution to integrating security among data handling. With proper incite, an intelligent Denial of Service (DoS) attack can still drastically effect an UWSN using information-centric components. In this work, we first analyze different forms of DoS attacks that take advantage of acoustic broadcast median and restricted modem energy. Next, we purpose a specialized algorithm to detect and cut off potential malicious nodes. Furthermore, we introduce machine learning techniques to evaluate detection rules overtime to better adhere to multiple mobile attackers. Simulation results depict the affects of this algorithm in the case of single and multiple DoS attacks. Our results show a strong correlation between DoS defensive methods and overall decreased network traffic.