P13ITE05

High Speed Networks

UNIT - I

Dr.A.Kathirvel
Professor & Head/IT - VCEW
 UNIT - I
 Frame Relay Networks
 Asynchronous transfer mode
 ATM protocol architecture
 ATM logical connection
 ATM cell and service categories – AAL
 High speed LANs: Fast, Gigabit ethernet, Fiber
channel
 Wireless LANs
 Introduction

 Packet-Switching Networks
 Switching Technique
 Routing
 X.25
 Frame Relay Networks
 Architecture
 User Data Transfer
 Call Control

3
 Packet-Switching Networks

 Basic technology the same as in the 1970s

 One of the few effective technologies for long
distance data communications
 Frame relay and ATM are variants of packet-
switching
 Advantages:
- flexibility, resource sharing, robust, responsive
 Disadvantages:
 Time delays in distributed network, overhead penalties
 Need for routing and congestion control

4
 Circuit-Switching

 Long-haul telecom network designed for voice

 Network resources dedicated to one call
 Shortcomings when used for data:
 Inefficient (high idle time)‫‏‬
 Constant data rate

5
 Packet-Switching

 Data transmitted in short blocks, or packets

 Packet length < 1000 octets
 Each packet contains user data plus control
info (routing)‫‏‬
 Store and forward

6
Figure 4.1 The Use of Packets

Chapter 4 Frame Relay 7

 Figure 4.2 Packet
Switching: Datagram
Approach

Chapter 4 Frame Relay 8

Advantages over Circuit-Switching

 Greater line efficiency (many packets can go

over shared link)‫‏‬
 Data rate conversions
 Non-blocking under heavy traffic (but
increased delays)‫‏‬

9
Disadvantages relative to Circuit-Switching

 Packets incur additional delay with every node

they pass through
 Jitter: variation in packet delay
 Data overhead in every packet for routing
information, etc
 Processing overhead for every packet at every
node traversed

10
Figure 4.3 Simple Switching Network

Chapter 4 Frame Relay 11

 Switching Technique

 Large messages broken up into smaller packets

 Datagram
 Each packet sent independently of the others
 No call setup
 More reliable (can route around failed nodes or
congestion)‫‏‬
 Virtual circuit
 Fixed route established before any packets sent
 No need for routing decision for each packet at
each node

12
Figure 4.4 Packet
Switching: Virtual-
Circuit Approach

Chapter 4 Frame Relay 13

 Routing

 Adaptive routing
 Node/trunk failure
 Congestion

14
 X.25

 3 levels
 Physical level (X.21)‫‏‬
 Link level (LAPB, a subset of HDLC)‫‏‬
 Packet level (provides virtual circuit
service)‫‏‬

15
Figure 4.5 The Use of Virtual Circuits

Chapter 4 Frame Relay 16

Figure 4.6 User Data and X.25
Protocol Control Information

Chapter 4 Frame Relay 17

 Frame Relay Networks

 Designed to eliminate much of the overhead in X.25

 Call control signaling on separate logical connection
from user data
 Multiplexing/switching of logical connections at layer
2 (not layer 3)‫‏‬
 No hop-by-hop flow control and error control
 Throughput an order of magnitude higher than X.25

18
Figure 4.7 Comparison of X.25 and
Frame Relay Protocol Stacks

Chapter 4 Frame Relay 19

Figure 4.8 Virtual Circuits and Frame
Relay Virtual Connections

Chapter 4 Frame Relay 20

 Frame Relay Architecture

 X.25 has 3 layers: physical, link, network

 Frame Relay has 2 layers: physical and data link (or
LAPF)‫‏‬
 LAPF core: minimal data link control
 Preservation of order for frames
 Small probability of frame loss
 LAPF control: additional data link or network layer
end-to-end functions

21
 LAPF Core

 Frame delimiting, alignment and transparency

 Frame multiplexing/demultiplexing
 Inspection of frame for length constraints
 Detection of transmission errors
 Congestion control

22
LAPF-core Formats

23
 User Data Transfer

 No control field, which is normally used for:

 Identify frame type (data or control)‫‏‬
 Sequence numbers
 Implication:
 Connection setup/teardown carried on separate
channel
 Cannot do flow and error control

24
 Frame Relay Call Control

 Frame Relay Call Control

 Data transfer involves:
 Establish logical connection and DLCI
 Exchange data frames
 Release logical connection

25
 Frame Relay Call Control

4 message types needed

 SETUP

 CONNECT

 RELEASE

 RELEASE COMPLETE

26
 ATM Protocol Architecture

 Fixed-size packets called cells

 Streamlined: minimal error and flow control
 2 protocol layers relate to ATM functions:
 Common layer providing packet transfers
 Service dependent ATM adaptation layer (AAL)‫‏‬
 AAL maps other protocols to ATM

27
 Protocol Model has 3 planes

 User
 Control
 management

28
29
 Logical Connections

 VCC (Virtual Channel Connection): a logical

connection analogous to virtual circuit in X.25

 VPC (Virtual Path Connection): a bundle of VCCs

with same endpoints

30
 Figure 5.2

Chapter 2 Protocols and the TCP/IP Suite 31

 Advantages of Virtual Paths

 Simplified network architecture

 Increased network performance and reliability
 Reduced processing and short connection setup time
 Enhanced network services

32
33
 VCC Uses

 Between end users

 Between an end user and a network entity
 Between 2 network entities

34
 Figure 5.3

Chapter 2 Protocols and the TCP/IP Suite 35

 VPC/VCC Characteristics

 Quality of Service (QoS)‫‏‬

 Switched and semi-permanent virtual channel
connections
 Cell sequence integrity
 Traffic parameter negotiation and usage monitoring
 (VPC only) virtual channel identifier restriction
within a VPC

36
 Control Signaling

 A mechanism to establish and release VPCs

and VCCs
 4 methods for VCCs:
 Semi-permanent VCCs
 Meta-signaling channel
 User-to-network signaling virtual channel
 User-to-user signaling virtual channel

37
 Control Signaling

 3 methods for VPCs

 Semi-permanent
 Customer controlled
 Network controlled

38
 ATM Cells

 Fixed size
 5-octet header
 48-octet information field
 Small cells reduce delay for high-priority cells
 Fixed size facilitate switching in hardware

39
 Header Format

 Generic flow control

 Virtual path identifier (VPI)‫‏‬
 Virtual channel identifier (VCI)‫‏‬
 Payload type
 Cell loss priority
 Header error control

40
 Figure 5.4

Chapter 2 Protocols and the TCP/IP Suite 41

 Generic Flow Control

 Control traffic flow at user-network interface (UNI)

to alleviate short-term overload conditions
 When GFC enabled at UNI, 2 procedures used:
 Uncontrolled transmission
 Controlled transmission

42
43
 Header Error Control

 8-bit field calculated based on remaining 32 bits of

header
 error detection
 in some cases, error correction of single-bit errors in
header
 2 modes:
 errordetection
 Error correction

44
 Figure 5.5

Chapter 2 Protocols and the TCP/IP Suite 45

 Figure 5.6

Chapter 2 Protocols and the TCP/IP Suite 46

 Figure 5.7

Chapter 2 Protocols and the TCP/IP Suite 47

 Service Categories

 Real-time service
 Constant bit rate (CBR)‫‏‬
 Real-time variable bit rate (rt-VBR)‫‏‬
 Non-real-time service
 Non-real-time variable bit rate (nrt-VBR)‫‏‬
 Available bit rate (ABR)‫‏‬
 Unspecified bit rate (UBR)‫‏‬
 Guaranteed frame rate (GFR)‫‏‬

48
 Figure 5.8

Chapter 2 Protocols and the TCP/IP Suite 49

 ATM Adaptation Layer (ATM)‫‏‬

 Support non-ATM protocols

 e.g., PCM voice, LAPF
 AAL Services
 Handle transmission errors
 Segmentation/reassembly (SAR)‫‏‬
 Handle lost and misinserted cell conditions
 Flow control and timing control

50
 Applications of AAL and ATM

 Circuit emulation (e.g., T-1 synchronous TDM

circuits)‫‏‬
 VBR voice and video
 General data services
 IP over ATM
 Multiprotocol encapsulation over ATM (MPOA)‫‏‬
 LAN emulation (LANE)‫‏‬

51
 AAL Protocols

 AAL layer has 2 sublayers:

 Convergence Sublayer (CS)‫‏‬
 Supports specific applications using AAL

 Segmentation and Reassembly Layer (SAR)‫‏‬

 Packages data from CS into cells and unpacks at

other end

52
 Figure 5.9

Chapter 2 Protocols and the TCP/IP Suite 53

 Figure 5.10

Chapter 2 Protocols and the TCP/IP Suite 54

 AAL Type 1

 Constant-bit-rate source
 SAR simply packs bits into cells and unpacks
them at destination
 One-octet header contains 3-bit SC field to
provide an 8-cell frame structure
 No CS PDU since CS sublayer primarily for
clocking and synchronization

55
 AAL Type 3/4

 May be connectionless or connection oriented

 May be message mode or streaming mode

56
57
 Figure 5.12

Chapter 2 Protocols and the TCP/IP Suite 58

 AAL Type 5

 Streamlined transport for connection oriented

protocols
 Reduce protocol processing overhead
 Reduce transmission overhead
 Ensure adaptability to existing transport protocols

59
 Figure 5.13

Chapter 2 Protocols and the TCP/IP Suite 60

61
 Emergence of High-Speed LANs

 2 Significant trends
 Computing power of PCs continues to grow
rapidly
 Network computing

 Examples of requirements
 Centralizedserver farms
 Power workgroups
 High-speed local backbone

62
 Classical Ethernet

 Bus topology LAN

 10 Mbps
 CSMA/CD medium access control protocol
 2 problems:
A transmission from any station can be received by
all stations
 How to regulate transmission

63
 Solution to First Problem

 Data transmitted in blocks called frames:

 User data
 Frame header containing unique address of
destination station

64
 Figure 6.1

Chapter 6 High-Speed LANs 65

 CSMA/CD

 Carrier Sense Multiple Access/ Carrier Detection

 If the medium is idle, transmit.

 If the medium is busy, continue to listen until the
channel is idle, then transmit immediately.
 If a collision is detected during transmission,
immediately cease transmitting.
 After a collision, wait a random amount of time, then
attempt to transmit again (repeat from step 1).

66
 Figure 6.2

Chapter 6 High-Speed LANs 67

Figure 6.3

Chapter 6 High-Speed LANs 68

 Medium Options at 10Mbps
 <data rate> <signaling method> <max length>
 10Base5
 10 Mbps
 50-ohm coaxial cable bus
 Maximum segment length 500 meters
 10Base-T
 Twisted pair, maximum length 100 meters
 Star topology (hub or multipoint repeater at central
point)‫‏‬

69
Figure 6.4

Chapter 6 High-Speed LANs 70

 Hubs and Switches

 Hub
 Transmission from a station received by central hub
and retransmitted on all outgoing lines
 Only one transmission at a time

 Layer 2 Switch
 Incoming frame switched to one outgoing line
 Many transmissions at same time

71
 Figure 6.5

Chapter 6 High-Speed LANs 72

Bridge Layer 2 Switch

 Frame handling done  Frame handling done

in software in hardware
 Analyze and forward  Multiple data paths
one frame at a time and can handle
 Store-and-forward multiple frames at a
time
 Can do cut-through

73
 Layer 2 Switches

 Flat address space

 Broadcast storm
 Only one path between any 2 devices

 Solution 1: subnetworks connected by routers

 Solution 2: layer 3 switching, packet-
forwarding logic in hardware

74
 Figure 6.6

Chapter 6 High-Speed LANs 75

Figure 6.7

Chapter 6 High-Speed LANs 76

Figure 6.8

Chapter 6 High-Speed LANs 77

 Figure 6.9

Chapter 6 High-Speed LANs 78

Figure 6.10

Chapter 6 High-Speed LANs 79

Figure 6.11

Chapter 6 High-Speed LANs 80

Benefits of 10 Gbps Ethernet over ATM

 No expensive, bandwidth consuming conversion

between Ethernet packets and ATM cells
 Network is Ethernet, end to end
 IP plus Ethernet offers QoS and traffic policing
capabilities approach that of ATM
 Wide variety of standard optical interfaces for 10
Gbps Ethernet

81
 Fibre Channel

 2 methods of communication with processor:

 I/Ochannel
 Network communications

 Fibre channel combines both

 Simplicity and speed of channel communications
 Flexibility and interconnectivity of network
communications

82
Figure 6.12

Chapter 6 High-Speed LANs 83

 I/O channel

 Hardware based, high-speed, short distance

 Direct point-to-point or multipoint communications
link
 Data type qualifiers for routing payload
 Link-level constructs for individual I/O operations
 Protocol specific specifications to support e.g.
SCSI

84
Fibre Channel Network-Oriented Facilities

 Full multiplexing between multiple destinations

 Peer-to-peer connectivity between any pair of ports
 Internetworking with other connection technologies

85
 Fibre Channel Requirements

 Full duplex links with 2 fibres/link

 100 Mbps – 800 Mbps
 Distances up to 10 km
 Small connectors
 high-capacity
 Greater connectivity than existing multidrop channels
 Broad availability
 Support for multiple cost/performance levels
 Support for multiple existing interface command sets

86
Figure 6.13

Chapter 6 High-Speed LANs 87

Fibre Channel Protocol Architecture

 FC-0 Physical Media

 FC-1 Transmission Protocol
 FC-2 Framing Protocol
 FC-3 Common Services
 FC-4 Mapping

88
 Wireless LAN Requirements
 Throughput
 Number of nodes
 Connection to backbone
 Service area
 Battery power consumption
 Transmission robustness and security
 Collocated network operation
 License-free operation
 Handoff/roaming
 Dynamic configuration
89
Figure 6.14

Chapter 6 High-Speed LANs 90

 IEEE 802.11 Services
 Association
 Reassociation
 Disassociation
 Authentication
 Privacy

91
Figure 6.15

Chapter 6 High-Speed LANs 92

 Figure 6.16

Chapter 6 High-Speed LANs 93

Questions ?

94