Anakyzing the performance of Active Queue Management Algorithms

      !

10.5121/ijcnc.2010.2201 1
   
       
  G.F.Ali Ahammed 1, Reshma Banu2, 1Department of Electronics & Communication, Ghousia college of Engg.Ramanagaram. ali_ahammed@rediffmail.com 2Department of Information Science & Engg, Ghousia college of Engg.Ramanagaram. ABSTRACT Congestion is an important issue which researchers focus on in the Transmission Control Protocol (TCP) network environment. To keep the stability of the whole network, congestion control algorithms have been extensively studied. Queue management method employed by the routers is one of the important issues in the congestion control study. Active queue management (AQM) has been proposed as a router-based mechanism for early detection of congestion inside the network. In this paper we analyzed several active queue management algorithms with respect to their abilities of maintaining high resource utilization, identifying and restricting disproportionate bandwidth usage, and their deployment complexity. We compare the performance of FRED, BLUE, SFB, and CHOKe based on simulation results, using RED and Drop Tail as the evaluation baseline. The characteristics of different algorithms are also discussed and compared. Simulation is done by using Network Simulator(NS2) and the graphs are drawn using X- graph. KEY WORDS RED; Droptail; Fairness Index; Throughput; AQM; NS2 FRED, BLUE, SFB, CHOKe, ECN

1. INTRODUCTION When there are too many coming packets contending for the limited shared resources, such as the queue buffer in the router and the outgoing bandwidth, congestion may happen in the data communication. During congestion, large amounts of packet experience delay or even be dropped due to the queue overflow. Severe congestion problems result in degradation of the throughput and large packet loss rate. Congestion will also decrease efficiency and reliability of the whole network, furthermore, if at very high traffic, performance collapses completely and almost no packets are delivered. As a result, many congestion control methods[2] are proposed to solve this problem and avoid the damage. Most of the congestion control algorithms are based on evaluating the network feedbacks[2] to detect when and where congestion occurs, and take actions to adjust the output source, such as reduce the congestion window (cwnd). Various feedbacks are used in the congestion detection and analysis. However, there are mainly two categories: explicit feedback and implicit feedback.
   In explicit feedback algorithms, some signal packets are sent back from the congestion point to warn the source to slow down [4], while in the implicit feedback algorithms, the source deduces the congestion existence by observing the change of some network factors, such as delay, throughput difference and packet loss [4]. Researchers and the IETF proposed active queue management (AQM) as a mechanism for detecting congestion inside the network.
             
          !

2 Further, they have strongly recommended the deployment of AQM in routers as a measure to preserve and improve WAN performance . AQM algorithms run on routers and detect incipient congestion by typically monitoring the instantaneous or average queue size. When the average queue size exceeds a certain threshold but is still less than the capacity of the queue, AQM algorithms infer congestion on the link and notify the end systems to back off by proactively dropping some of the packets arriving at a router. Alternately, instead of dropping a packet, AQM algorithms can also set a specific bit in the header of that packet and forward that packet toward the receiver after congestion has been inferred. Upon receiving that packet, the receiver in turns sets another bit in its next ACK.
When the sender receives this ACK, it reduces it transmission rate as if its packet were lost. The process of setting a specific bit in the packet header by AQM algorithms and forwarding the packet is also called marking. A packet that has this specific bit turned on is called a marked packet. End systems that experience the marked or dropped packets reduce their transmission rates to relieve congestion and prevent the queue from overflowing. In practice, most of the routers being deployed use simplistic Drop Tail algorithm, which is simple to implement with minimal computation overhead, but provides unsatisfactory performance. To attack this problem, many queue management algorithms are proposed, such as Random Early Drop (RED) [3], Flow Random Early Drop (FRED) [4], BLUE [5], Stochastic Fair BLUE (SFB) [5], and CHOKe (CHOose and Keep for responsive flows, CHOose and Kill for unres

…(Full text truncated)…