tcp-vegas.h

Go to the documentation of this file.

/*
 * Copyright (c) 2016 ResiliNets, ITTC, University of Kansas
 *
 * This program is free software; you can redistribute it and/or modify
 * it under the terms of the GNU General Public License version 2 as
 * published by the Free Software Foundation;
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program; if not, write to the Free Software
 * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA  02111-1307  USA
 *
 * Author: Truc Anh N. Nguyen <annguyen@ittc.ku.edu>
 *
 * James P.G. Sterbenz <jpgs@ittc.ku.edu>, director
 * ResiliNets Research Group  https://resilinets.org/
 * Information and Telecommunication Technology Center (ITTC)
 * and Department of Electrical Engineering and Computer Science
 * The University of Kansas Lawrence, KS USA.
 */
 
#ifndef TCPVEGAS_H
#define TCPVEGAS_H
 
#include "tcp-congestion-ops.h"
 
namespace ns3
{
 
class TcpSocketState;
 
/**
 * \ingroup congestionOps
 *
 * \brief An implementation of TCP Vegas
 *
 * TCP Vegas is a pure delay-based congestion control algorithm implementing a proactive
 * scheme that tries to prevent packet drops by maintaining a small backlog at the
 * bottleneck queue.
 *
 * Vegas continuously measures the actual throughput a connection achieves as shown in
 * Equation (1) and compares it with the expected throughput calculated in Equation (2).
 * The difference between these 2 sending rates in Equation (3) reflects the amount of
 * extra packets being queued at the bottleneck.
 *
 *              actual = cwnd / RTT        (1)
 *              expected = cwnd / BaseRTT  (2)
 *              diff = expected - actual   (3)
 *
 * To avoid congestion, Vegas linearly increases/decreases its congestion window to ensure
 * the diff value fall between the 2 predefined thresholds, alpha and beta.
 * diff and another threshold, gamma, are used to determine when Vegas should change from
 * its slow-start mode to linear increase/decrease mode.
 *
 * Following the implementation of Vegas in Linux, we use 2, 4, and 1 as the default values
 * of alpha, beta, and gamma, respectively.
 *
 * More information: http://dx.doi.org/10.1109/49.464716
 */
 
class TcpVegas : public TcpNewReno
{
  public:
    /**
     * \brief Get the type ID.
     * \return the object TypeId
     */
    static TypeId GetTypeId();
 
    /**
     * Create an unbound tcp socket.
     */
    TcpVegas();
 
    /**
     * \brief Copy constructor
     * \param sock the object to copy
     */
    TcpVegas(const TcpVegas& sock);
    ~TcpVegas() override;
 
    std::string GetName() const override;
 
    /**
     * \brief Compute RTTs needed to execute Vegas algorithm
     *
     * The function filters RTT samples from the last RTT to find
     * the current smallest propagation delay + queueing delay (minRtt).
     * We take the minimum to avoid the effects of delayed ACKs.
     *
     * The function also min-filters all RTT measurements seen to find the
     * propagation delay (baseRtt).
     *
     * \param tcb internal congestion state
     * \param segmentsAcked count of segments ACKed
     * \param rtt last RTT
     *
     */
    void PktsAcked(Ptr<TcpSocketState> tcb, uint32_t segmentsAcked, const Time& rtt) override;
 
    /**
     * \brief Enable/disable Vegas algorithm depending on the congestion state
     *
     * We only start a Vegas cycle when we are in normal congestion state (CA_OPEN state).
     *
     * \param tcb internal congestion state
     * \param newState new congestion state to which the TCP is going to switch
     */
    void CongestionStateSet(Ptr<TcpSocketState> tcb,
                            const TcpSocketState::TcpCongState_t newState) override;
 
    /**
     * \brief Adjust cwnd following Vegas linear increase/decrease algorithm
     *
     * \param tcb internal congestion state
     * \param segmentsAcked count of segments ACKed
     */
    void IncreaseWindow(Ptr<TcpSocketState> tcb, uint32_t segmentsAcked) override;
 
    /**
     * \brief Get slow start threshold following Vegas principle
     *
     * \param tcb internal congestion state
     * \param bytesInFlight bytes in flight
     *
     * \return the slow start threshold value
     */
    uint32_t GetSsThresh(Ptr<const TcpSocketState> tcb, uint32_t bytesInFlight) override;
 
    Ptr<TcpCongestionOps> Fork() override;
 
  protected:
  private:
    /**
     * \brief Enable Vegas algorithm to start taking Vegas samples
     *
     * Vegas algorithm is enabled in the following situations:
     * 1. at the establishment of a connection
     * 2. after an RTO
     * 3. after fast recovery
     * 4. when an idle connection is restarted
     *
     * \param tcb internal congestion state
     */
    void EnableVegas(Ptr<TcpSocketState> tcb);
 
    /**
     * \brief Stop taking Vegas samples
     */
    void DisableVegas();
 
  private:
    uint32_t m_alpha;             //!< Alpha threshold, lower bound of packets in network
    uint32_t m_beta;              //!< Beta threshold, upper bound of packets in network
    uint32_t m_gamma;             //!< Gamma threshold, limit on increase
    Time m_baseRtt;               //!< Minimum of all Vegas RTT measurements seen during connection
    Time m_minRtt;                //!< Minimum of all RTT measurements within last RTT
    uint32_t m_cntRtt;            //!< Number of RTT measurements during last RTT
    bool m_doingVegasNow;         //!< If true, do Vegas for this RTT
    SequenceNumber32 m_begSndNxt; //!< Right edge during last RTT
};
 
} // namespace ns3
 
#endif // TCPVEGAS_H