PI for one of the sub research topics about microservice platform management, Key R & D plan of the Ministry of science and technology,￥13 million，I am responsible for ~￥38,2500，2021.1-2023.12
PI, SIAT Innovation Program for Excellent Young Researchers, Supported by SIAT-CAS, ￥80,000, 2020.1-2021.12
Co-PI, CloudSim Research Project for Siemens MindSphere, Supported by Siemens (Chengdu) Company,￥150,000, 2019.9-2019.12
Part of a team for Software Defined Approach and Key Technology of Smart Management of Cloud Data Centers, Key-Area Research and Development Program of Guangdong Province (No.2020B010164003), ￥20 million, 2020.1-2022.12
Mr. Qiheng Zhou (2019.7-2020.6, undergraduate from SYSU, one CCGRID 2020 paper, one SPE 2021 paper, →master at NUS)
Mr. Lingxiao Xu (2020.2-2020.12, postgraduate from UESTC, one CollaborateCom 2020 paper)
Mr. Lei Yang (2020.12-2021.6, MSc at SIAT, co-supervised with Prof. Yang Wang, → Alibaba Inc.)
Mr. Zhiheng Zhong (2021.4-now, PhD at UniMelb)
Mr. Cundao Yu (2021.6-now, BSc at UESTC)
Projects with codes:
CoScal: Deep Learning based Approach for Cloud Workloads Prediction
Predicted Alibaba and Google cloud workloads based on supervised learning and gradient recurrent units.
Reinforcement learning based approach for scaling decision, including vertical scaling, horizontal scaling and brownout mechanisms.
Energy Efficient Algorithms based on VM Consolidation Evaluated in CloudSim (CCGRID 2020)
Implemented algorithms published in TPDS, JPDC, TSC, TCC and etc.
A software prototype system implementing brownout based on containers. The testbed is based is based on Grid'5000. Containers are managed by Docker Swarm. The containers are controlled to save energy consumption.
heduling process, we propose FlexCloud, a new flexible and scalable simulator that eBrownout is an approach enables to dynamically deactivate/acticate optional application components to handle overloads. We have implemented brownout techniques in CloudSim to reduce energy consumption in data centers.
February 2013 to June 2013
Cloud Data centers aim to provide reliable, sustainable and scalable services for different requests. Resource scheduling is a key part for Cloud services. To simplify the scheduling process, we propose FlexCloud, a new flexible and scalable simulator that enables simulating the process of initializing cloud data centers, allocating virtual machine requests, and providing performance evaluation for various scheduling algorithms.