Leading Prof.

Prof. Ke XU (徐恪教授), Department of Computer Science & Technology, Tsinghua University.


  • NewJune 2021, Congratulations! Our paper is awarded Best Paper by IEEE/ACM IWQoS.
  • (New)May 2021, Congratulations! A new book (高级计算机网络(第2版)) is published. 
  • NewMarch 2021, Congratulations! Our paper is accepted by ACM CCS.​​
  • NewJanuary 2021, Congratulations! Our paper is accepted by IEEE Network.
  • NewJuly 2020, Congratulations! Our paper is accepted by ACM CCS.​​
  • NewMay 2020, Congratulations! Our paper is accepted by ACM SIGCOMM.​​
  • Congratulations! A new book (算法统治世界 智能经济的隐形秩序) is published.
  • For more news, please click here

Call for papers

  • Asiacrypt ’20, TH-CPL A. Deadline: May 22, 2020, link
  • ACM IMC’21, TH-CPL A. Deadline: May 26, 2020, link
  • IEEE S&P’21, TH-CPL A. Summer deadline: June 4, 2020, link
  • ACM CoNEXT’20, TH-CPL A. Deadline: June 17, 2020, link
  • NDSS’21, TH-CPL A. Fall deadline: July 17, 2020, link
  • NSDI’21,  TH-CPL A. Fall deadline: September 10, 2020​, link
  • TIFS, TH-CPL A. Special Issue on Signal Processing Advances in Wireless Transmission of Information and Power, Deadline: November 30, 2020, link
  • TDSC, TH-CPL A. Special Issue on Software Reliability and Dependability Engineering, Deadline: March 16, 2021, link

Research achievements

1. Research on Evolvable Internet Architecture There is a general consensus about the success of Internet architecture in academia and industry. However, with the development of diversified application, the existing Internet architecture is facing more and more challenges in scalability, security, mobility and performance aspects. We proposed a novel evolvable Internet architecture framework to meet the continuous changing application requirements. The basic idea of evolvability is relaxing the constraints that limit the development of the architecture while adhering to the core design principles of the Internet. Three important design constraints used to ensure the construction of the evolvable architecture, including the evolvability constraint, the economic adaptability constraint and the manageability constraint, are comprehensively described. We consider that the evolvable architecture can be developed from the network layer under these design constraints. What’s more, we believe that the address system is the foundation of the Internet.  Therefore, we propose a general address platform which provides a more open and efficient network environment for the research and development of the evolvable architecture. 2. Research on the Scalable and Reconfigurable High-performance Routing and Switching Technology This research presents a BGP implementing model based on the scalable routing and switching System, which has a higher speedup ratio compared with that of the previous proposals. It proposes the function module division and system development method at three levels – sub-component, component and macro-component – based on the characteristic of routing and switching software. Compared with other methods, this one raises the code reuse proportion and efficiently reduces the routing system development cost. From this research, six patents of invention have been granted. The scalable and reconfigurable high-performance router software platform has better compatibility, and provides developing and testing tools with higher performance efficiency compared with Cisco IOS system and MIT Click platform, respectively. The result helped win the Science and Technology of Electronic Information Award by the Chinese Institute of Electronics (first class) in 2012. This software platform has been applied to routers of Bitway Networking Technology Co., Ltd., and adopted by network operators, such as CERNET2, China Telecom, and China Mobile, realizing desirable economic and social achievements. Many communication companies have also adopted this software when developing products or testing performance, including H3C, Ruijie Networks and Maipu Communication Technology Practical application by these companies shows “the application of this platform greatly improves development efficiency and accelerates research progress.” 3. Research on P2P Network Traffic Classification & Management and Live Streaming Technology This research proposes a new P2P network traffic estimation model, which possesses higher estimation accuracy than the recently proposed gravity and independent connection model. A P2P network traffic classification method is proposed based on P2P networks’ transmission behavior. This method has a simpler approach to classification with a higher success rate than the best result of previous classification methods proposed based on the Bayes’ Theorem. The paper has been cited for ten times since it was published in Computer Communications. A cache replacement algorithm based on P2P network live streaming media is proposed; it has higher cache efficiency than the previous representative cache method based on segments. A node selection and a transmission scheduling method based on P2P network live streaming media are proposed, solving the problem that the traditional random selection method can hardly guarantee the transmission quality of streaming media and being granted two patents of invention. Based on this achievement, Prof. Xu’s group and PPLive cooperated to develop and implement the cross-platform P2P video accelerator, PPVA. According to the measurement and statistics of the PPVA website, there have been 280 million times of downloading this software, which supports thousands of video web sites invisibly with more than 10 million on-line users downloading videos every day, effectively saving the network broad-band of video web sites and network operators. 4. Research on Internet Traffic Management Network multipath traffic management and optimal traffic engineering are studied based on Network Utility Maximization (NUM) theory, which leads to the multipath traffic management method based on the logarithmic barrier method. The multipath traffic management method can realize theory optimization and convergence assurance simultaneously that the previous representatives, including TRUMP, can’t achieve. A new load balancing in proportion method based on traffic engineering is proposed, getting a new conclusion that the optimal routing in load balancing in proportion is the routing with the shortest M/M/1 queuing delay. Based on this conclusion, the optimal traffic engineering can be realized through routing protocols designed to find the shortest path, such as OSPF, in network engineering.

Tsinghua University
Network Technology Research Institute of Tsinghua University