Wenyu Ren
  • HOME
  • EDUCATION
  • RESEARCH
    • Online, Context-aware, Intelligent Anomaly Detection and Analysis for SCADA Systems
    • Intelligent Synchrophasor Data Real-Time Compression
    • Applied Resiliency for More Trustworthy Grid Operation (ARMORE)
    • Secure and Scalable Data Collection Protocol in Smart Grid
    • Optimal Data Replication in RSU Aided Vehicles Networks
    • Limits of Predictability and Patterns of Vehicular Mobility
    • Optimal Vehicle Number and Coding Decision in Vehicular DTNs
    • Multi-Armed Bandit Problem
    • Green Data Center
  • PUBLICATION
  • ABOUT ME
  • CV

Green Data Center

Advisor: Associate Prof. Dan Li
Computer Network Lab in THU
March 2011–August 2011
Brief Introduction
Today’s giant data centers are power hungry. Data center energy saving not only controls the operational cost, but also benefits the environments. IT equipments dominate the energy consumption in modern data centers and it is increasingly important to reduce the energy consumed by the network part of IT equipments. This project aims at designing energy efficient flow scheduling mechanism to reduce the network energy footprint of distributed computation in container-sized data centers. A central controller periodically schedules the flows to accommodate flow dynamics, network dynamics and inaccurate estimation. Using Fat-Tree as a representative data center network architecture, we exploit the topological characteristic to develop an online flow scheduling algorithm. We evaluate the topology-aware flow scheduling by real MapReduce workloads from Terasort and Wordcount applications. Simulation results show that the algorithm can save 15%~40% network energy compared with the ECMP flow scheduling. It also outperforms the classical optimization algorithms such as simulated annealing and particle swarm optimization, in terms of both the network energy saving and the scheduling speed.
My Contribution
I studied  the topology  in  data  center  and  heuristic  algorithms including simulated  annealing  and particle swarm. I also implemented  our  own  topology-aware flow scheduling  algorithm  simulation  using  C++  and  made comparisons with other  heuristic  algorithms. 
Powered by Create your own unique website with customizable templates.