Current research areas:

• "Autonomic Cloud Resource Management"

  While Cloud Computing systems are demanded highly, their management is getting a serious concern, especially when the power and performance constraints are considered. The aim of this project is to provide an Autonomic Management service to reduce the power consumption of the cloud computing systems while maintaining the required performance. We have demonstrated that by first analyzing and characterizing the workload behavior on such systems, right amount of resources can be allocated that provide the required performance with upto 84% less energy consumption.

• "Autonomic Resilient Cloud Services"

  For the cloud computing paradigm to be fully adopted and effectively used, it is critical that the security mechanisms are robust and resilient to faults and attacks since cloud system security is extremely complex due to the many interdependent tasks and it is well accepted that building a cloud computing system without any attacks is not practical. The concern grows especially when the mission critical systems are considered. As a solution, we provide resiliency based on Software Behavior Obfuscation, Moving Target Defense (MTD), redundancy, and Autonomic Computing.

Former research areas:

• "Computer Arithmetic" with Prof. H. Fatih Ugurdag at an Embedded and MicroElectronic Systems Lab (nEMESysLab)

  Built a framework for benchmarking various summation and multiplication methods. Also, worked on a novel algorithm that has novelties compared to Wallace and Dadda by employing a technique called plowing in order to reduce the final adder.

• "Variation and Defect Tolerance for Nano Crossbars" with Prof. Mehdi Tahoori at Dependable Nano Computing Lab (DNL)

  Issues in CMOS technology are getting more challenging and as an alternative, nano crossbars with a PLA-like behavior are considered. However, variations and defects are two major problems due to low controllability of manufacturing and should be tolerated during the design phase. The goal of the project was to develop both variation and defect tolerant logic mapping algorithms for nano crossbars that could be used for design automation. Additionally, characterization testing method for crossbars has been proposed.