Startup founder, computer networks researcher, mobile app engineer, experience in big data and machine learning
My early research focuses on improving mobile network performance. I designed and developed measurement mechanisms to measure Wi-Fi and LTE performance and energy consumption on mobile devices. The measurement data showed that existing LTE control algorithms cause significant energy waste which results in short battery life for mobile devices. Thus, I developed a new LTE control algorithm that maintains network performance while reducing energy consumption by up to 75%. We also observed significant temporal and spatial unbalanced performance for Wi-Fi and LTE on a single mobile device, which causes current multi-interface network protocols to run at sub-optimal performance. To optimize network performance, I then developed a machine learning based algorithm to dynamically allocate network traffic on both interfaces. This algorithm improves network throughput by up to 90%.
I have worked on a cloud computing measurement and job placement study on Amazon EC2 and Rackspace datacenters, which provides guidelines for datacenter users to allocate large-scale, network intensive workloads in a performance-orientated manner.