Research
I am a Ph.d. student at the University of Central Florida. I have completed my course work for the Ph.D. program and am actively engaged in research now. I currently work in the field of Evolvable Hardware. Our research group is currently developing the Competitive Runtime Reconfiguration (CRR) approach to effect refurbishment of FPGAs at runtime in Xilinx FPGAs.
Our approach provides a method to identify the presence of faults in FPGAs using the idea of competition among configurations. Over a period of time, using output comparisons with the normal data-flow inputs, we can identify the faulty resources, as well as affected configurations. The health-states of configurations are monitored, and using an embedded evolutionary algorithm, we seek to repair the affected configurations. The health of the configurations is restored, and these refurbished configurations are then viable for normal use. The various characteristics such as throughput, fault-repair latency and correctness are variable according to application-specific needs. We hope to demonstrate a system that can autonomously detect faults and maintain availability while effecting refurbishment of the device, without taking the device offline. The project is sponsored by NASA.
My work involves formalizing the methods, deriving performance bounds, improving the isolation, and comprehensive analysis and modeling of the system. Research is thrilling, and very involving - going from a concept, or an idea, to an implementation, and then understanding how it works is very interesting.
Besides evolvable hardware, I sustain an interest in cryptography, fault tolerant computing, and computer system design in general.