Chapter 10 Congestion Control in Data Networks

1

Congestion Control in Data Networks and Internets

COMP5416Chapter 10

Chapter 10 Congestion Control in Data Networks 2

Review Performance and QoS are key design

requirements for networks Greater PC computing power, distributed

applications, multimedia contents driven the needs for higher capacity LANs (i.e. Gigabit Ethernet) and WANs (ATM & WDM)

Key to design is ability to model and estimate performance metrics

Has profound effects on network configurations and protocol design

Queueing analysis and simulations are some tools Key to monitor networks in (near) congestion:

Need to device congestion and traffic management tools

Chapter 10 Congestion Control in Data Networks 3

Introduction Congestion occurs when number of

packets transmitted approaches network capacity

Objective of congestion control: keep number of packets below level

at which performance drops off dramatically

Chapter 10 Congestion Control in Data Networks 4

Queuing Theory Data network is a network of

queues If arrival rate > transmission rate

(>) queue size grows without bound

and packet delay goes to infinity ()

Chapter 10 Congestion Control in Data Networks 5

Chapter 10 Congestion Control in Data Networks 6

At Saturation Point, 2 Strategies

Discard any incoming packet if no buffer available

Saturated node exercises flow control over neighbors May cause congestion to propagate

throughout network

Chapter 10 Congestion Control in Data Networks 7

Chapter 10 Congestion Control in Data Networks 8

Ideal Network Performance I.e., infinite buffers and no overhead for

packet transmission or congestion control Throughput increases with offered load

until full capacity Packet delay increases with offered load

approaching infinity at full capacity Power = throughput / delay Higher throughput results in higher delay

Chapter 10 Congestion Control in Data Networks 9

Chapter 10 Congestion Control in Data Networks 10

Practical Performance I.e., finite buffers and non-zero packet

processing overhead With no congestion control, increased

load eventually causes moderate congestion: throughput increases at slower rate than load

Further increased load causes packet delays to increase and eventually throughput to drop to zero

Chapter 10 Congestion Control in Data Networks 11

Chapter 10 Congestion Control in Data Networks 12

Congestion Control Approaches

Backpressure Request from destination to source to

reduce rate Choke packet: ICMP Source Quench

Implicit congestion signalling Source detects congestion from

transmission delays and discarded packets and reduces flow

Chapter 10 Congestion Control in Data Networks 13

Explicit congestion signaling Direction

Backward Forward

Categories Binary Credit-based rate-based

Chapter 10 Congestion Control in Data Networks 14

Chapter 10 Congestion Control in Data Networks 15

Traffic Management Issues Fairness

Last-in-first-discarded (i.e. drop-tail) may not be fair

Quality of Service (QoS) – provision of service differentiation Voice, video: delay sensitive, loss insensitive File transfer, mail: delay insensitive, loss

sensitive Interactive computing: delay and loss sensitive

Reservations Policing: excess traffic discarded or handled on

best-effort basis

TM CC

Chapter 10 Congestion Control in Data Networks 16

Example: Frame Relay A high-performance WAN protocol that

operates at the physical and data link layers of the OSI reference model

It provides connection-oriented link layer communication Connection exists between each pair of

devices and are associated with a connection identifier (DLCI)

Two categories of virtual connections: switched virtual circuits (SVCs) permanent virtual circuits (PVCs)

Chapter 10 Congestion Control in Data Networks 17

Frame Relay Stack

Chapter 10 Congestion Control in Data Networks 18

Frame Relay Congestion Control Implements a simple congestion-notification

mechanisms (i.e. binary) rather than explicit, per-virtual-circuit flow control

Flow control left to higher-layer protocols FR uses two congestion-notification mechanisms:

Forward-explicit congestion notification (FECN) Backward-explicit congestion notification (BECN)

Each is controlled by a one bit in FR frame header Header also contains a Discard Eligibility (DE) bit

which is used to identify less important frames that can be dropped during periods of congestion

Chapter 10 Congestion Control in Data Networks 19

2 Bits for Explicit Signaling Forward Explicit Congestion Notification

For traffic in same direction as received frame

This frame has encountered congestion Backward Explicit Congestion Notification

For traffic in opposite direction of received frame

Frames transmitted may encounter congestion

Chapter 10 Congestion Control in Data Networks 20

Chapter 10 Congestion Control in Data Networks 21

Traffic Rate Management Committed Information Rate (CIR)

Rate that network agrees to support Aggregate of CIRs < capacity

For node and user-network interface Committed Burst Size (Bc)

Maximum data over one interval agreed to by network

Excess Burst Size (Be) Maximum data over one interval that

network will attempt

Chapter 10 Congestion Control in Data Networks 22

Figure 10.6

Chapter 10 Congestion Control in Data Networks 23

Figure 10.7

BC – committed burst size

Be – excess burst size

Chapter 10 Congestion Control in Data Networks 24

Summary Congestion control and traffic

management are required to ensure acceptable network performance

Frame relay has simple schemes to handle traffic

Next: TCP traffic management & control

Traffic Mgmt

CongestionControl