My research interests include -





My research work experience -

Assessment of WiMAX Technology for performance, Interoperability and Manageability on Campus Area Test Bed, I.I.T. Delhi (June 2007 - July 2008) - Supervised by Prof. Huzur Saran : I worked as a research associate on this project under Prof. Huzur Saran and Prof. B.N.Jain in the Advanced Network Lab. My team was responsible for the establishment of a campus-wide WiMAX testbed (pre-WiMAX and 802.16d 802.16e standards), followed by its performance assessment and analysis. My involvement in this project was with the establishment of the campus area testbed for the pre-WiMAX and 802.16d standard and its performance analysis and assessment. I was also responsible for network monitoring of the equipment thus installed using JRuby scripts.

Mobile Device Fingerprinting - Device ID in Enterprise Networks (March 2010 - December 2013) - Supervised by Prof. Huzur Saran : Modern heterogeneous networks implement two-factor authentication. Identification at the client end is typically performed by software-based parameters (cookies, OS, browser, etc). In case of change of hardware for device, the device identification remains the same thereby leaving the user data vulnerable. In my research, I provide a systematic evaluation of the stability of clock skew based device fingerprint for hardware device identification. I perform a multi-dimensional (measurement methods, measurement environments and configurations) comparison of factors for all device kinds in a heterogeneous network. I provide optimizations for the initial introduced approach. I discover handheld skew jumps (device manufacturer dependent) under varying operating power and ambient temperatures. Thus, this research provides an empirical study to understand the full-scale capabilities of skew fingerprint in a moderate-size heterogeneous network, to aid the network administrator/NIDS/NIDS.

Repeatability and Verifiability in Networking Experiments, Information Sciences Institute, University of Southern California, Los Angeles, Calfornia, USA (July 2011 - December 2014) - Supervised by Dr. Alefiya Hussain : Networking experiments are stochastic and possess many sources of variation due to different experiment components - deterministic (base code), dynamic (background network behavior) and possible opportunistic (cyber-attack). These experiments are repeated on different network topologies, experimentation platforms and hardware apparatus at different time periods by independent researchers. Thus, it is very difficult to verify an experiment's execution. In my research, I propose a direct way to characterize an experiment execution. I introduce a framework that facilitates comparison and subsequent verification of repeated experiment executions based on event logs and measurement data. Experiment signatures (Time-Series, Markov Chain based) are compared via standard distance measures (KL divergence, Total Variation Distance and Euclidean Distance). I demonstrate that my framework is statistically rigorous, sound and sensitive to changes in experiment configuration (hardware, topology, traffic and mobility). I extensively test this framework on a few experiment classes (opportunistic and heavy-tailed) and on different experimentation platforms (simulators, emulated testbeds and real-networks). This research is my first step in generating a validity management framework for determining repeatability in experiment executions. To the best of my knowledge this research is the first in this direction and possess great potential.