Active Queue Management (AQM) based Internet Congestion Control October 1 2002 Seungwan Ryu (sryu@eng.buffalo.edu) PhD Student of IE Department University at Buffalo
Contents Internet Congestion Control Active Queue Management (AQM) Control-Theoretic design of AQM Performance Evaluation Summary and Issues for Further study References
I. Internet Congestion Control Internet Traffic Engineering What is Congestion? Congestion Control and Avoidance TCP Congestion Control Active Queue management (AQM) Other Approaches
Internet Traffic Engineering Measurement: for reality check Experiment: for Implementation Issues Analysis: Bring fundamental understanding of systems May loose important facts because of simplification Simulation: Complementary to analysis: Correctness, exploring complicate model May share similar model to analysis
What is Congestion? What is congestion? The aggregate demand for bandwidth exceeds the available capacity of a link. What will be occur? Performance Degradation Multiple packet losses Low link utilization (low Throughput) High queueing delay Congestion collapse
What is congestion? (2) Congestion Control Open-loop control Mainly used in circuit switched network (GMPLS) Closed-loop control Mainly used in packet switched network Use feedback information: global & local Implicit feedback control End-to-end congestion control Examples: TCP Tahoe, TCP Reno, TCP Vegas, etc. Explicit feedback control Network-assisted congestion control Examples: IBMSNA,DECbit,ATMABR,ICMP source quench, RED, ECN
Congestion Control and Avoidance Two approaches of handling Congestion Congestion Control (Reactive) Play after the network is overloaded Congestion Avoidance (Proactive) Play before the network becomes overloaded
Paradigms of the Current Internet Paradigms: For design and Operation: Keep it simple Design principle of TCP: Do not ask the network to do what you can do yourself These paradigms are aimed for best-effort service As the Internet evolves and grows in size and number of users, the network has experienced performance degradation such as more packet drop In addition, service evolves to a variety of services Question: Do we need new paradigm?
TCP Congestion Control Uses end-to-end congestion control Uses implicit feedback e.g., time-out, triple duplicated ACKs, etc. Uses window based flow control cwnd = min (pipe size, rwnd) self-clocking slow-start and congestion avoidance Examples: TCP Tahoe, TCP Reno, TCP Vegas, etc.
TCP Congestion Control (2) Slow-start and Congestion Avoidance cwnd W* Slow Start Congestion Avoidance W+1 W*/2 RTT RTT Time
TCP Congestion Control (3) TCP Tahoe Use slow start/congestion avoidance Fast retransmit: an enhancement detect packet (segments) drop by three duplicate ACKs W = W/2, and enter congestion avoidance TCP Reno (fast recovery) Upon receiving three duplicate ACKs ssthresh = W/2, and retransmit missing packets Upon receiving next ACK: W = ssthresh Allow the window size grow fast to keep the pipeline full
TCP Congestion Control (4) TCP SACK (Selected Acknowledgement) TCP (Tahoe) sender can only know about a single lost per RTT SACK option provides better recovery from multiple losses The sender can transmit all lost packets But those packets may have already been received Operation Add SACK option into TCP header The receiver sends back SACK to sender to inform the reception of the packet Then, the sender can retransmit only the missing packet
Other Approaches : Pricing Smart-market [Mackie-Mason 1995] A price is set for each packet depends on the level of demand for bandwidth Admit packets with bid prices that exceed the cut-off value The cut-off is determined by the marginal cost Paris metro pricing (PMP) [Odlyzko] To provide differentiated services The network is partitioned into several logical separate channels with different prices With less traffic in channel with high price, better QoS would be provided.
Other approaches (2): Optimization Concept Network resource allocation problem: User problems Network problems User problem sends bandwidth request with price Network problem allocate bandwidth to each users by solving NLP User problem Users can be distinguished by a utility function A user wants to maximize its benefit (utility - cost) Network problem maximize aggregate utilities subject to the link capacity constraints Then, it can be formulated to a Non-linear programming (NLP) problem
Other approaches (3): Fairness Two fairness issues Fair bandwidth sharing: network-centric Fair packet drop (mark): user-centric Fair bandwidth sharing Max-min fair [Bertsekas, 1992]: No rate can be increased without simultaneous decreasing other rate which is already small provides equal treatment to all flows Proportional fair [Kelly 1998] A feasible set of rates are non-negative and the aggregate rate is not greater than link capacity and the aggregate of proportional change is zero or negative provides different treatment of each flow according to their rates
II. Active Queue Management (AQM) Internet Congestion Control Active Queue Management (AQM) Control-Theoretic design of AQM Performance Evaluation Summary and Issues for Further study References
Active Queue Management (AQM) What is AQM? Examples of AQM: RED and Variants More about AQM: Extensions
Active Queue Management (AQM) Performance degradation in current TCP Congestion Control Multiple packet loss Low link utilization Congestion collapse The role of the router becomes important Control congestion effectively in networks Allocate bandwidth fairly
AQM (2) Problems with current router algorithm Use FIFO based tail-drop (TD) queue management Two drawbacks with TD: lock-out, full-queue Lock-out: a small number of flows monopolize usage of buffer capacity Full-queue: The buffer is always full (high queueing delay) Possible solution: AQM Definition: A group of FIFO based queue management mechanisms to support end-to-end congestion control in the Internet
AQM (3) Goals of AQM Reducing the average queue length: Decreasing end-to-end delay Reducing packet losses: More efficient resource allocation Methods: Drop packets before buffer becomes full Use (exponentially weighted) average queue length as an congestion indicator Examples: RED, BLUE, ARED, SRED, FRED,.
AQM (4) Random Early Detection (RED) Use network algorithm to detect incipient congestion Design goals: minimize packet loss and queueing delay avoid global synchronization maintain high link utilization removing bias against bursty source Achieve goals by randomized packet drop queue length averaging
RED avgq = (1 WQ ) avgq + W Q Q P P d = p max 0 avg min Q th maxth min 1 th min avg th Q avg max th < min Q th < max avg Q th
AQM (5) : BLUE Algorithm Upon packet loss if (now - last_update >freeze_t) Pm = pm + d1 last_update = now upon link idle if (now - last_update >freeze_t) Pm = pm - d2 last_update = now Concept To avoid drawbacks of RED Parameter tuning problem Actual queue length fluctuation Decouple congestion control from queue length Use only loss and idle event as an indicator Maintains a single drop prob., p m Drawback Can not avoid some degree of multiple packet loss and/or low utilization
Algorithm AQM (6) : SRED i th arriving packet is compared with a randomly selected one from Zombie list Hit = 1, if they are from same flow =0,ifNOT p(i)=hit frequency=(1-α)p(i-1)+αhit p(i) -1 : estimator of # of active flows Packet drop probability 1 Pzap = Psred min(1, (256 P( i)) p sred = (1/ 4) pmax (1/ 6) B q < (1/ 3) B 0 q < (1/ 6) B p * 2 max ) (1/ 3) B q < B Concept stabilize queue occupancy use actual queue length Penalize misbehaving flows Drawbacks P(i) -1 is not a good estimator for heterogeneous traffic Parameter tuning problem: P sred, P zap,etc. Stabilize queue occupancy when traffic load is high. What happen when traffic load is low?
AQM (7) : ARED Adapt aggresiveness of RED according to the traffic load change adapt max p based on queue behavior Operation Increase max p when avg Q crosses above max th Decrease max p when avg Q crosses below min th freeze max p after changing to prevent oscillation
More about AQM Responsive (TCP) vs. unresponsive flows (UDP) RED fail to regulate unresponsive flows UDP do not adjust sending rate upon receiving congestion signal UDP flows consumes more bandwidth than fair share FRED [Lin & Morris, 1997] Tracks the # of packets in the queue from each flow maintain logical queues for each active flows in a FIFO queue Fair share for a flow is calculated dynamically Unresponsive flows are identified and penalized Drop packets proportional to bandwidth usage See TCP-friendly website (http://www.psc.edu/networking/tcp_friendly.html)
More about AQM (2) Supporting QoS and DiffServ with AQM Try to support a multitude of transport protocol (TCP, UDP, etc.) Classify several types of services rather than one best-effort service. Then, apply different AQM control to each services classes. Examples: RIO (RED In and Out) [Clark98] CBT (Class based Thresholds) [Floyd1995]
More about AQM (3) RIO (RED in and out) [Clark 1998] Separate flows into two classes: IN and OUT service profile Router maintains two different statistics for each service profiles. Different parameters and average queue lengths Avgs: forin packet: avg IN, for OUT profile: avg TOTAL When congested, apply different control to each classes Drop Prob. 1 P max_out P max_in Min th_out Max th_out Max th_in avg =Min th_in
More about AQM (4) CBT [Floyd 1995] packets are classified into several classes maintain a single queue but allocate fraction of capacity to each class Apply AQM (RED) based control to each class Once a class occupies its capacity, discard all arriving packets Drawbacks Fairness problem in case of changing traffic mix static threshold setting Total utilization can be fluctuated Dynamic-CBT [Chung2000] Track the number of active flows of each class dynamically adjust threshold values of each class