Conference Paper

CHOKeW: bandwidth differentiation and TCP protection in core networks

Dept. of Electr. & Comput. Eng., Florida Univ., Gainesville, FL
DOI: 10.1109/MILCOM.2005.1605882 Conference: Military Communications Conference, 2005. MILCOM 2005. IEEE
Source: IEEE Xplore


Both bandwidth differentiation and TCP protection is important for implementing quality of services (QoS) in TCP/IP networks. To the best of our knowledge, no other schemes have combined these two tasks together so far. In this paper we present a stateless active queue management (AQM) algorithm called CHOKeW. CHOKeW uses "matched drops" created by CHOKe to control the bandwidth allocation, but excludes the RED module. Based on the congestion status and the priority of the arriving packet, CHOKeW adjusts the maximum number of packets drawn from the buffer for matched drops. Using simulations, we show that CHOKeW is capable of working in different congestion scenarios, supporting multiple bandwidth priority levels by giving high priority flows with high throughput, and restrict the throughput of high-speed unresponsive flows to protect TCP flows

Download full-text


Available from: Yuguang Fang
  • Source
    • "Previous works [5] and [14] have shown that the approximate sending rate of TCP can be described as (10) where denotes the combination of packet size, roundtrip time, and other parameters such as the TCP version and the speed of user's computers. The sending rate of TCP decreases as networking congestion deteriorate [19], [20], so we have "
    [Show abstract] [Hide abstract]
    ABSTRACT: Although differentiated services (DiffServ) networks have been well discussed in the past several years, a conventional Active Queue Management (AQM) algorithm still cannot provide low-complexity and cost-effective differentiated bandwidth allocation in DiffServ. In this paper, a novel AQM scheme called CHOKeR is designed to protect TCP flows effectively. We adopt a method from CHOKeW to draw multiple packets randomly from the output buffer. CHOKeR enhances the drawing factor by using a multistep increase and single-step decrease (MISD) mechanism. In order to explain the features of CHOKeR, an analytical model is used, followed by extensive simulations to evaluate the performance of CHOKeR. The analytical model and simulation results demonstrate that CHOKeR achieves proportional bandwidth allocation between different priority levels, fairness guarantee among equal priority flows, and protection of TCP against high-speed unresponsive flows when network congestion occurs.
    Preview · Article · Feb 2014 · IEEE Transactions on Industrial Informatics
  • [Show abstract] [Hide abstract]
    ABSTRACT: In order to find a cost-effective solution for combining pricing, admission control, buffer management and scheduling in DiffServ networks, we propose the ConTax-CHOKeW framework. ConTax is a distributed admission controller that works in edge routers, while CHOKeW, designed in our previous research for core networks, is a buffer management scheme. Since a user in a higher priority class requests more bandwidth that is differentiated by CHOKeW, this user is likely to generate more network load. ConTax uses the sum of priority values for all admitted users to measure the network load. When the network load is heavier, ConTax charges a higher price for each priority class. The price includes a basic price and an extra price called congestion tax. By using simulations, we show that ConTax is capable of controlling the network load as well as the number of users that are admitted into the network. In addition, by employing ConTax, network providers can gain more profit, and users have greater flexibility that in turn meets the specific performance requirements of their applications
    No preview · Article · Oct 2006
  • [Show abstract] [Hide abstract]
    ABSTRACT: The Internet is gradually and constantly becoming a multimedia network that needs mechanisms to provide effective quality of service (QoS) requirements to users. The service curve (SC) is an efficient description of QoS and the service curve based earliest deadline first policy (SCED) is a scheduling algorithm to guarantee SCs specified by users. In SCED, deadline calculation is the core. However, not every SC has a treatable deadline calculation; currently the only known treatable SC is the concave piecewise linear SC (CPLSC). In this paper, we propose an algorithm to translate all kinds of SCs into CPLSCs. In this way, the whole Internet can have improved performance. Moreover, a modification of the deadline calculation of the original SCED is developed to obtain neat and precise results. The results combining with our proposed algorithm can make the deadline calculation smooth and the multimedia Internet possible.
    No preview · Article · Jun 2008 · Information Sciences
Show more