Student Spotlight: Nidhi Madabhushi's Thesis
Authored by Nidhi Madabhushi
My interest in doing research and subsequently a thesis started almost three years ago, when I was in my senior year, pursuing Bachelor’s in Computer Science in Oman. It was during this time in 2017 that I started studying about Blockchain Technology, IoT and Network Security. I submitted four papers to different conferences and also ended up winning the best paper award for one of them. This motivated me and solidified my intention for further study in the area of security and I was thrilled when I was accepted to the Master’s in Computer Science program at the University of Denver in Fall 2018.
I studied in different countries – India, The Sultanate of Oman and now the USA and I think the resources and environment that the Ritchie School offers here at DU is really world-class. I was lucky to get a tour of the ECS building on my first day at DU and Meredith Corley was very kind to walk me around and answer a lot of my questions. As I settled into my course work, I realized that the choice of a thesis topic was not an easy one to make, and so, I took some time during my first year at DU to explore a couple different areas before I finally found interest in anomaly detection. This was a topic that was introduced in the network security course taught by Professor Rinku Dewri, who was also my advisor. His approach of teaching was intriguing, with discussions around current research problems that are unsolved in that area. After many meetings and reading various research papers, I started to work on my thesis titled ‘Real-time Detection of Demand Manipulation Attacks on a Power Grid’. This research is about providing techniques to detect attacks that use high wattage IoT devices to disrupt the power grid. I was also interested in creating a project that would positively impact and contribute to the research community. With these as my driving goals for my thesis, we evaluated different existing models and proposed a novel technique to effectively detect such kind of attacks in a power grid.
As with any research, it was not a smooth ride in the beginning, and I experienced a steep learning curve. I tried various implementations and techniques and most of them failed from a technical and functional standpoint. However, Professor Rinku was very patient, and he always encouraged me throughout the entire research experience. We also set up specific timelines and broke down the goals to a weekly level to make it more manageable. I was also able to remotely access and use the systems in the cybersecurity lab to run my lengthy time-consuming scripts and collect the results effectively. Great infrastructure at Ritchie School that provided me uninterrupted access to the systems I needed and regular and structured communication with my advisor, ensured my research was not affected by the COVID-19 situation. With many helpful inputs from Dr. Rinku Dewri, the technical part of my thesis picked up pace and neared completion.
Once that was done, I began to put the research into writing and started to gather all the concepts implemented. Since this work was done over many months, I found it very challenging to bring together all the contributions and connect the dots in order to present the bigger conceptual picture. I enjoyed writing the results section of my thesis as I started finding new insights into it and this helped me prepare for my defense as well. One of the tips I got from my advisor that worked really well was to present the work like an interesting story rather than reading out boring bullet points. He also recommended to add in appealing visualizations and design a power grid application that simulates the attacks and shows how our method detects them. After practicing and preparing well for the presentation, I felt confident and successfully defended my thesis. I would like to thank all the defense committee members for their time and support. It felt very rewarding in the end, especially after all the hard work that was put into the thesis over the two years during my master’s program. I am immensely grateful to my advisor Dr. Dewri for the valuable guidance and to the Ritchie School for all the facilities. My biggest learning from this experience is to make small and consistent progress each day to get to the final goal. Another important takeaway is to practice well for the defense at least a week in advance, as it is not easy present about two years’ worth of extensive research within 50 minutes.
My research experience had many ups and downs, but I enjoyed it as every time something did not work, it gave me an opportunity to learn and build something new. This motivated me to continue my academic journey and start a PhD in Computer Science at DU from Winter 2021 to further this research with Dr. Rinku Dewri as my advisor. I hope my experience helps graduate students follow their research dreams with the help of the encouraging faculty and facilities available here at the Ritchie School.