Jiaqiang Liu (柳嘉强)  

Ph.D Candidate, Dept. of Electronic Engineering, Tsinghua University

+86-133-3113-2364
liujiaqiang12@mails.tsinghua.edu.cn
Rohm Building 10-202, Tsinghua University, Beijing, 100084, China.

I am a fourth year Ph.D student in the Department of Electronic Engineering of Tsinghua University, co-advised by Prof. Depeng Jin and Prof. Yong Li. I received the B.E. degree from the same department in 2012.

My researches mainly focus on the application of software-defined network and network function virtualization in the mobile core network and data center network, including the re-design of the overall network architecture and the resource allocation and scheduling problems in the new architecture. I also have interests in big data driven network optimization. I have participated in multiple joint research projects with Huawei, Hitachi and ZTE.

Here is my CV.


Publications


  • Jiaqiang Liu, Yong Li, Ying Zhang, Li Su, Depeng Jin. Improve Service Chaining Performance with Optimized Middlebox Placement, In IEEE Transactions on Service Computing, DOI 10.1109/TSC.2015.2502252. .
  • Jiaqiang Liu, Yong Li, Huandong Wang, Depeng Jin, Li Su, Lieguang Zeng, Thanos Vasilakos. Leveraging Software-Defined Networking for Security Policy Enforcement, In Information Science, Vol. 327, 2016, pp: 288–299..
  • Jiaqiang Liu, Yong Li, Depeng Jin, Lieguang Zeng. Traffic-Aware Cross-Site Virtual Machine Migration in Future Mobile Cloud Computing, In Mobile Networks and Applications, Vol. 20. No. 1, 2015, pp. 62—71. .
  • Jiaqiang Liu, Yong Li, Min Chen, Wenxia Dong, Depeng Jin. Software-Defined Internet of Things for Smart Urban Sensing, In Communication Magazine. Vol. 53, Issue. 9, 2015, pp: 55-63. .
  • Huan Yan, Jiaqiang Liu, Yong Li, Wenxia Dong, Chengyong Lin, Depeng Jin. WAN as a Service for Cloud via Software-Defined Network and Open APIs, In Proc. of INFOCOM 2015. .
  • Jiaqiang Liu, Li Su, Yuchen Jin, Yong Li, Depeng Jin. Optimal VM Migration Planning for Data Centers, In Proc. of Globecom 2014. .
  • Jiaqiang Liu, Yong Li, Depeng Jin. SDN-based Live VM Migration Across Datacenters, In Proc. of SIGCOMM 2014. .
  • Jiaqiang Liu, Depeng Jin. Software Defined Live Virtual Machine Migration, In Proc. of ICNP 2013. .
  • Jiaqiang Liu, Yong Li, Yin Shi, Juan Wu, Depeng Jin. A Method and System to Measure VM Dependencies. Patent. Application No. 201510100366.0. .

Professional Activities

  • Teaching Assistant for Undergraduate Course: Theories and Experiments of Computer Networks, 2014-2016.
  • Conference Paper Reviewer of GLOBECOM 2015, ICCCN 2015
  • Journal Paper Reviewer of IEEE Communication Magazine, Mobile Networks and Applications, Frontiers of Computer Science.

Projects

SDN and NFV based Mobile Core Network

Currently, the mobile core network are implemented by dedicated hardware appliances, which are cost expensive and hard to manage. This project, cooporated with Huawei, aims to re-design the mobile core network by using SDN and NFV. Specifically, we propose to totally separate the control plane and user plane functions of mobile core network, and provide them via virtual machines in the cloud. The control plane is logically centralized, and dynamically instantiates instances of user plane functions according to traffic load and service requirements. We study the design of the overall network architecture and the specific way for separating the control and user plane functions of DPI, caching and charging. We implement a prototype system using OpenStack and OpenDaylight.

Discovering and Understanding Viewing Video Patterns in the Urban Neighbourhood

Video occupies a large proportion of traffic on the Internet. Understanding its geographical viewing patterns is extremely valuable for the design of Internet ecosystems. This project aims to investigate the problem that whether there exists distinct viewing patterns among the neighborhoods of an urban city. We design a system that utilizes LDA model and K-Means clustering to automatically extract geographical video viewing patterns from viewing video requests. We make a rigorous measurement study of two months real video requests in Shanghai, collected from one major ISP of China. We illustrate that there exists four prevalent and distinct video access patterns, and they are corresponding to four different geographical context: downtown resident, office regions, suburb resident and hybrid regions.

Software Defined Data Center Network

Today, a IT company usually has multiple geographicallly distrbuted data centers to improve user experience and service robustness. In this project, we propose a two layer architecture for the control of distributed data center network. Based on this architecture we study the VM migration procedure and the mechanisms to optimize it. We implement the two layer control architecture and software-defined VM migration procedure using Floodlight and OpenStack.

End-to-end Network Performance Measurement for the Cloud

Accurate and comprehensive end-to-end network performance measurement is critical for the Cloud for the automatic trouble shooting and optimized provision of various services. In this project, we design NetWatch, a system that provides measurement as a service through open APIs. We address the security, multi-path forwarding and multi-tenant issues to enable NetWatch to provide accurate and low-overhead measurement services. We implement NetWatch in OpenStack Neutron and show its application and performance under various scenarios.