Introduction to ATM
Handout
Released
Short Title ATM All rights reserved. Passing on and copying of this document, use and communication of its contents not permitted without written authorization from Alcatel.
2 / 106
Contents
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Why a B-ISDN ? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1 The N-ISDN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1.1 The N-ISDN is Narrow . . . . . . . . . . . . . . 1.1.2 The N-ISDN is Half-Integrated . . . . . . . 1.1.3 The N-ISDN is 64 Kbit/sec - Based . . . Market Pull & Technology Push . . . . . . . . . . . . . . . . . . Requirements for the B-ISDN . . . . . . . . . . . . . . . . . . . Evolution of Transfer Modes . . . . . . . . . . . . . . . . . . . . 2.1.1 Evolution from Circuit Switching . . . . . . . 2.1.2 Evolution from Packet Switching . . . . . . . . The Asynchronous Aspect . . . . . . . . . . . . . . . . . . . . . . No Link-by-link Error Control . . . . . . . . . . . . . . . . . Connection Oriented . . . . . . . . . . . . . . . . . . . . . . . . . . Small, Fixed Header and Information Field Sizes . . . Transparent Delivery of Information . . . . . . . . . . . . . . The Header Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . GFC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VPI / VCI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PTI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CLP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . HEC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.6.1 Header Error Protection . . . . . . . . . . . . . . . 4.6.2 Cell delineation . . . . . . . . . . . . . . . . . . . . . The Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2.1 The Physical Layer . . . . . . . . . . . . . . . . . . . 5.2.2 The ATM Layer . . . . . . . . . . . . . . . . . . . . . . 5.2.3 The AAL . . . . . . . . . . . . . . . . . . . . . . . . . . . . The User Plane and the Control Plane . . . . . . . . . . . . 9 11 12 12 12 14 16 18 21 22 23 26 29 30 33 34 35 37 39 40 41 42 46 47 48 48 50 53 54 54 55 56 57 59 61
The ATM Cell Header . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1 4.2 4.3 4.4 4.5 4.6
5.3 6
Call Configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3 / 106
Contents
Multi-Connection Calls . . . . . . . . . . . . . . . . . . . . . . . . Point-to-Multipoint Calls . . . . . . . . . . . . . . . . . . . . . . ATM Traffic Control Functions . . . . . . . . . . . . . . . . . . . 7.1.1 CAC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.2 UPC / NPC . . . . . . . . . . . . . . . . . . . . . . . . . 7.1.3 ABR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Protocol Stacks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2.1 DSS2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.2.2 B-ISUP . . . . . . . . . . . . . . . . . . . . . . . . . . . . PDH versus SDH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ATM Mapping in PDH . . . . . . . . . . . . . . . . . . . . . . . . . 9.2.1 Mapping into Lower Order Frames . . . . . 9.2.2 Mapping into E3 Frames . . . . . . . . . . . . . . ATM Mapping in SDH . . . . . . . . . . . . . . . . . . . . . . . . . IMA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
4 / 106
Contents
Figures
Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Figure 13 Figure 14 Figure 15 Figure 16 Figure 17 Figure 18 Figure 19 Figure 20 Figure 21 Figure 22 Figure 23 Figure 24 Figure 25 Figure 26 Figure 27 Figure 28 Figure 29 Figure 30 Figure 31 Figure 32 Figure 33 Figure 34 Figure 35 Figure 36 Figure 37 Figure 38 Figure 39 Figure 40 Figure 41 Figure 42 Figure 43 Figure 44 Integrated Access of the N-ISDN . . . . . . . . . . . . . . . . . . . . . . Non-Integrated Core Network of the N-ISDN . . . . . . . . . . The B-ISDN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Technology Push and Market Pull. . . . . . . . . . . . . . . . . . . . . . . A B-ISDN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Spectrum of Transfer Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . Circuit Switching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Packet Switching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Packets and Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Processing Functionalities on different layers. . . . . . . . . . . . . ATM Cell Stream . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Different Transport Strategies . . . . . . . . . . . . . . . . . . . . . . . . . . STM versus ATM . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Packets transmitted without Link-by-Link Error Control . . . Packets transmitted with Link-by-Link Error Control . . . . . . CO & CL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . An ATM Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The UNI and NNI Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . The ATM Cell Header . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . VPI & VCIs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A VP Switch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A VP / VC Switch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Local Significance of VPI / VCI . . . . . . . . . . . . . . . . . . . . . . . . . The PTI bits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The CLP bit . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The HEC Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Cell Delineation Algorithm . . . . . . . . . . . . . . . . . . . . . . . . The B-ISDN Protocol Reference Model . . . . . . . . . . . . . . . . . Different Physical Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Different AALs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The AAL-5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Control Plane and User Plane for a Switch . . . . . . . . . . . . . . . A Unidirectional Point-to-Multipoint Connection . . . . . . . . CAC functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Traffic Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UPC and NPC Locations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ideal UPC / NPC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conforming and non-conforming Cells . . . . . . . . . . . . . . . . The GCRA (T, t ) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Feedback Loop of RM cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . Feedback Loop of RM cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . Signalling Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Protocol Stack for DSS2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Protocol Stack for B-ISUP using an ATM network . . . . . . . . . , 12 13 14 17 19 22 24 26 27 28 30 31 32 33 33 34 36 40 40 42 43 44 45 46 47 49 51 54 56 57 58 59 63 66 67 69 70 70 71 72 74 76 78 79
5 / 106
Contents
Figure 45 Figure 46 Figure 47 Figure 48 Figure 49 Figure 50 Figure 51 Figure 52 Figure 53 Figure 54 Figure 55 Figure 56 Figure 57 Figure 58
Protocol Stack for B-ISUP reusing NB links . . . . . . . . . . . . . . , PDH and SDH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PCM Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Back to Back Multiplexing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Add/Drop Multiplexing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ring Shaped SDH Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . PDH versus SDH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mapping of ATM Cells into E1 . . . . . . . . . . . . . . . . . . . . . . . . . E3 frame for ATM Cells. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mapping of ATM Cells into E3 . . . . . . . . . . . . . . . . . . . . . . . . . STM-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mapping of ATM Cells into a VC-4 . . . . . . . . . . . . . . . . . . . . Inverse muxing, cell by cel. . . . . . . . . . . . . . . . . . . . . . . . . . . . . IMA Frames . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
80 82 83 84 85 85 87 88 89 90 91 92 94 95
6 / 106
Contents
Tables
Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 Table 10 Table 11 Advantages of CO & CL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Advantages Fixed or Variable Cells . . . . . . . . . . . . . . . . . . . . . Advantages of Small and Large ATM Cells . . . . . . . . . . . . . . . Proposals for ATM Payload and Header Lengths . . . . . . . . . . ATM Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ABR Binary Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Signalling Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PDH Multiplexing Hierarchies . . . . . . . . . . . . . . . . . . . . . . . . . . SDH Containers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The SDH Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . PDH versus SDH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 35 35 36 55 73 76 83 86 86 87
7 / 106
Contents
8 / 106
Preface
Preface
This document is to be used during an ATM overview training.
9 / 106
Preface
10 / 106
1 Why a B-ISDN ?
1 Why a B-ISDN ?
This chapter describes why and how the concept of a new network, called a Broadband Integrated Services Digital Network (B-ISDN), came into existence. Essentially, there are 3 critiques one can give regarding the current Narrowband ISDN network. These will be examined and will lead to a set of requirements for the B-ISDN. Finally, this chapter describes the terms of technology push" and market pull".
11 / 106
1 Why a B-ISDN ?
1.1
1.1.1
The N-ISDN
The N-ISDN is Narrow
In the N-ISDN, the order of magnitude of bitrates is about a few Kbit/sec. In the B-ISDN, the order of magnitude of bitrates is about a few Mbit/sec.
1.1.2
User Perspective
On one hand, the user perceives the network as providing services in a uniform (i.e. an integrated) manner.
Figure 1
Network Perspective
On the other hand, once inside the network, things are different. Dedicated networks handle different kinds of data..Examples of such dedicated networks are :
"
A Packet Switching Network : to support interactive data traffic in a cost-effective way. This network processes data messages in a packetised form. A Frame Relay Network : high speed data network, which interconnects local area data networks. Neither the telephone- nor the packet switching network can offer the same speed of processing. A Video Conference Network : offers video conferencing facilities on a national or international level. A Cable Television Network : supports television and radio distribution functions.
"
"
"
12 / 106
1 Why a B-ISDN ?
PSTN
Data
Internet
Figure 2
Half-Integration
This evolution leads to a multitude of networks, each capable to support a limited number of functions. This approach has the following disadvantages :
"
"
Interworking problems :
D D
"
Inefficient Solution :
D
Multi-Service Network
To avoid these disadvantages, the idea emerged to create a single, high capacity network, which is able to flexibly support all types of services (voice, data, video). Services required by the present and by the future users; consequently, this new network also has to be future safe.
13 / 106
1 Why a B-ISDN ?
At the same time it also simplifies the user access. As a result, this universal network leads to substantial cost savings and to a better infrastructure for all users. The name given to this network is Integrated Broadband Communication Network (IBCN) or B-ISDN.
B-ISDN
Figure 3
The B-ISDN
1.1.3
Voice Frequencies are 300 to 3400 Hz. Double sampling (Nyquest) : 8000 Hz. Each sample is 1 byte long = 8 bits. 8 bits * 8000 /sec = 64 Kbit/sec
For this reason, as well transmission technology and equipment and switching equipment are generally designed to transfer 64Kbit/sec.. However, nowadays, new services come up. For example :
"
Modern communication is not limited anymore to communication between human beings. Computers have to communicate with each other too. They send data information over networks.
14 / 106
1 Why a B-ISDN ?
"
Video-telephony is a voice and video service. Users experience it as a telephone service, but with the additional possibility to be able to see the other communicating party. The video conference service offers the opportunity to people to take part on meetings without having to travel to meet each other. This service can offer serious cost savings for the participants. The Video On Demand (VOD) service offers an easy and comfortable way to watch video movies. It is an alternative to video rental shops. Television broadcast quality can be improved through the usage of digital technology. At the same time a new type of television is being developed, High Definition Television (HDTV) Multimedia services comprise any combination of voice, data and video services.
"
"
"
"
These new services do not at all require a rate of 64 Kbit/sec. Therefore, we can distinguish 2 basic problems.
Not all intermediate values of bitrates can be reached. What must you allocate for a variable bandwidth source ?
" "
The highest bitrate ? Then you over-allocate. The average bitrate ? Then you risk quality reduction.
15 / 106
1 Why a B-ISDN ?
1.2
Technology Push
This (r)evolutionary process can be approached from the side of the technology. The development of the technology can offer new, until now unimaginable possibilities to the users. So, technological developments open new horizons, stimulating the definition of new services. This is referred to as the "technology push ". In recent years, large progress was achieved both in the field of electronics and in the field of optical technology. Among the technological achievements the following can be mentioned:
"
Semi-conductor technology: CMOS technology allows to construct high complexity, high speed systems on a very small chip surface. Optical technology: advances in optical transmission offer higher speed and lower costs. Three parameters of the transmission are constantly being enhanced: the bit rate, the distance, and the quality. Video coding algorithms: the advances made in the video coding algorithms allow the more efficient distribution of the bandwidth among different simultaneous applications. Less bandwidth is required per user while ensuring the same quality of picture. ...
"
"
"
Market Pull
Nevertheless, even if new technology makes new things possible, users and in this case must be willing to invest money, in other words, is there a demand from the market. This is referred to as the "market pull ". It means that there is a need for a B-ISDN, because of the changing and more demanding user requirements.
16 / 106
1 Why a B-ISDN ?
B-ISDN
Actual Situation Figure 4 Technology Push and Market Pull. The market pull can only become strong enough because of the rise of a certain Killer Application". Possible candidates for this were (or are ) :
" "
17 / 106
1 Why a B-ISDN ?
1.3
Service independent network. Interactive and Distributive services are transported over the network: Both interactive, point-to-point services like telephone calls, as well as distributive services like multicasting are handled. High Bitrate, Broadband network: The network has a large capacity, it offers its services to many users and many applications. Typical bitrates of the network are 155 Mbit/s, 622 Mbit/s, or even more. The network is capable to offer bandwidth usage on demand. It is suitable to carry information for Fixed Bit Rate services (of any bitrate) and for Variable Bit Rate services. Future Safe network: The network has to be able to adapt itself to changes in technology and user requirements. In the future new services can arise, which are unknown today. These services will all have to be treated by that same network. Exchanges Transmission links Access networks
"
"
"
ATM
To realise such a B-ISDN network a new technology is necessary, which describes how the information has to be treated in that network. This technogly is Asynchronous Transfer Mode (ATM) and it requires a new type of switches, called ATM switches. Usually, the interconnections of the exchanges are realised by a digital transmission system, which is called Synchronous Digital Hierarchy (SDH)
SDH
18 / 106
1 Why a B-ISDN ?
LAN
SDH Links
Figure 5
A B-ISDN
This document is about ATM, SDH, PDH and their relationship. These technologies are a means to provide a new, multi-service, high-bandwidth, future-safe network.
19 / 106
1 Why a B-ISDN ?
20 / 106
21 / 106
2.1
In what follows, the possible paths starting from these transfer modes are examined. Figure 6 shows a proposal for a spectrum" of the different transfer modes.
Unflexible and Easy Make Flexible Flexible and Complex Simplify
Frame Relaying
ATM
Figure 6
22 / 106
2.1.1
When the multiplexed information stream enters the telephone exchange, a circuit switch switches every timeslot to a certain destination. The incoming information is identified by the incoming physical link and timeslot number, the outgoing information is identified by the outgoing physical link and timeslot number. This is shown on figure 7. The circuit switch takes his routing decisions depending upon a routing table. The entries in this table can be filled in, because of
" "
23 / 106
Incoming links
4 3 2 1 4
Outgoing links
3 2 1
I1 I2 In time
O1 O2 Om time
Routing Table
Incoming link
Timeslot
Outgoing link
Timeslot
Connection
I1 I1 I1 I2
1 2 4 4
O1 O2 Om O2
2 1 2 4
Figure 7
Circuit Switching With the transfer mode of circuit switching as starting point, we can try to increase its flexibility.
Multirate Switching
One improvement is create a timeslot structure, existing of timeslots of different sizes. In this way the network is able to support a small set of different bitrates. However, we are still faced with the already mentioned problems :
" "
Further, creating a timeslot structure of different sizes of timeslots has the additional problem of inefficiency.
24 / 106
A second improvement is the technique of fast circuit switching. In this transfer mode, a user can request as many timeslots, as it instantaneously needs, but this amount does not need to stay fixed for the complete duration of the connection. Therefore, the user will have to use fast signalling to allocate or deallocate timeslots has he needs them. This solutions seems to meet all the requirements for a B-ISDN, but is has one crucial disadvantage :
"
Complexity
25 / 106
2.1.2
Incoming links B A I1 I2 C In O1 O2 Om E
Outgoing links D
time
time
Routing Table Incoming link Incoming Header A B C Outgoing link Outgoing Header E F D
I1 I1 In
O2 Om O1
Figure 8
Packet Switching When data packets arrive in a switch, error control is performed. To correct an erroneous data packet, retransmission of this information is necessary. Meanwhile the right sequence of the data packets that belong to the same communication also has to be maintained. This means, that a high amount of processing is
26 / 106
required for every packet, and this in every switch within the network. This complexity results in delays and thus in a low throughput capability of the packet switching network. If we examine this complexity more closely, we see that a lot of processing is done on different layers, i.e.
"
On the layers 2, exchanging frames, the functionalities of flow control and error control are performed.
"
Flag
Address
Info
CRC
Packet
27 / 106
Layer 3
Flow Control
Layer 3
Layer 2
Layer 2
Layer 1 Figure 10
Layer 1
The first improvement is to omit flow control on layer 3 This technique is called frame switching The second improvement is to omit flow control on layer 2 This technique is called frame relaying The final improvement is to omit error control on layer 2 This technique comes close to ATM
"
"
Currently, it is not anymore a problem to omit all this functionality in the network, because the quality of the networks increases
Conclusion
The moral of this story is that ATM combines the best of both circuit switching and packet switching, i.e. it combines
" "
28 / 106
29 / 106
3.1
cell
time
Figure 11
Asynchronous
Let us be precise about the exact meaning of the term Asynchronous" in ATM :
"
No No Yes
"
"
30 / 106
Analogy
Figure 12
Figure 13 gives a comparison between circuit switching (or Synchronous Transfer Mode (STM)) and ATM.
31 / 106
Channel 1 Channel 2
.. . Channel n - Reservation of Physical Channels - Fixed number of channels. - Fixed bitrate of channels. - Low granularity of bitrates. - Physical resources are permanently reserved, even if nothing is transmitted. Figure 13 STM versus ATM
- Reservation of Virtual Channels - Variable number of channels. - Variable bitrates of channels. - High granularity of bitrates. - Physical resources are not used, if nothing is transmitted.
Disadvantages of ATM
Overhead Complexity
32 / 106
3.2
Figure 14
Figure 15
In general, the effect of omitting link-by-link error control is traffic increase. The higher the Bit Error Rate (BER) of the link, the higher this effect will be. Therefore, for ATM to be a suitable transfer mode, the BER of the links must be lower than 10-5. Note that, even if we omit link-by-link error control, this does not mean that no error control can be done anymore. An application can still implement end-to-end error control. The point is that the network will not be responsible for this anymore, in other words, the payload information is completely transparent to the network.
33 / 106
3.3
Connection Oriented
A general can instruct an army convoy to get to a destination in different ways :
" "
B
Figure 16 CO & CL Of course, both methods offer advantages : Table 1 Advantages of CO & CL Advantages of CL No Setup Time Easy Reaction to Failures Load Distribution
34 / 106
3.4
"
Table 2
Advantages Fixed or Variable Cells Advantages of Variable Cell Length No Half-Empty Cells
Advantages of Fixed Cell Length Easy for Buffers, Memory, Processing Delay
"
Table 3
Advantages of Small and Large ATM Cells Advantages of Large Cell Length Efficiency
35 / 106
Table 4
Proposals for ATM Payload and Header Lengths Payload Length Header Length 6 4 5
64 32 48
Thus, all ATM cells consists of 53 bytes, 5 bytes of header and 48 bytes of payload.
Payload
Header
48 Bytes
5 Bytes
Figure 17
An ATM Cell
36 / 106
3.5
37 / 106
38 / 106
39 / 106
4.1
UNI
NNI
Figure 18
Header Formats
as is shown in figure . The difference between both formats is only in the first 4 bits.
UNI
8 7 6 5 4 3 2 1 1 2 8 7 6 5
NNI
4 3 2 1
VPI
VPI
PTI HEC
CLP
4 5
Figure 19
Generic Flow Control (GFC) Virtual Path Identifier (VPI) Virtual Cell Identifier (VCI) Payload Type Identifier (PTI) Cell Loss Priority (CLP) Header Error Control (HEC)
40 / 106
UNI
4.2
GFC
The GFC mechanism assists in the control of the flow of traffic from ATM connections . More specifically, the GFC mechanism is used to control traffic flow in order to alleviate short-term overload conditions that may occur. GFC is only available at the user-to-network interface. Any end user-to-end user flow control requires additional network facilities.
41 / 106
4.3
VPI / VCI
2 levels
In the ATM network, every virtual connections is identified by the fields VPI and VCI. These fields provide a 2-level identification. AnTwo types of identifiers are defined:
" "
VC VC VC
VP
VP
VC VC VC
VC VC VC
VP
Physical Medium
VP
VC VC VC
VC VC VC
VP
VP
VC VC VC
Figure 20
VPI & VCIs To work with this 2-level identification, offers some advantages :
"
Easy Global Referring (VP Switching) It is very common to do this. (cfr. Street + Number, First Name + Last Name)
"
"
42 / 106
VP Switch
A switch that only examines the VPI field is called a VP switch. It does not interpret nor change the value of the VCI. It performs routing, based upon the VPI only and VPI translation.
Incoming links
VCI 23 VCI 45
VPI 1
VCI 67
VPI 2 Om time
VPI 4
VCI 23 VCI 45
time
VPI 1 VPI 2
VPI 4 VPI 3
Figure 21
A VP Switch
43 / 106
VP / VC Switch
A switch that examines both the VPI and the VCI fields is called a VP / VC switch. It performs routing, based upon the VPI and the VCI, and does VPI / VCI translation. It is said a VP / VC switch terminates VPs.
Incoming links
VCI 23 VCI 45
VPI 1
VPI 2 Om
VCI 67
VPI 4
VCI 23 VCI 45
time
VPI in
VCI in
VPI out
VCI out
Figure 22
A VP / VC Switch
44 / 106
Local Significance
In the ATM network, virtual connections are established between the communicating users. Each virtual connection is identified by a number, which has only local, link-by-link significance in the virtual connection. Explicitely, the VPI / VCI field does not designate an end address. Compared with circuit switching, it is more like a timeslot identifier". Thus, the VPI / VCI field of an ATM cell will be modified as it traverses several ATM switches.
Figure 23
Local Significance of VPI / VCI Which VPI / VCI will be used on which link, will be agreed upon by the exchanges during the setup of the ATM connection. This way of working has the following advantage :
"
45 / 106
4.4
PTI
This parameter identifies more or less the type of the payload". The 3 bits all have a different meaning.
PTI
0 1
0 1 No Congestion Congestion
Figure 24
46 / 106
4.5
CLP
The CLP bit labels an ATM cell with a certain priority, as shown in figure 25.
CLP
X
0 High Priority 1 Low Priority Figure 25 The CLP bit This priority can be important in case of congestion Therefore, the .Cell Loss Ratio (CLR) will be different for these different priorities : CLR1 > > CLR0 A cell priority can be assigned :
" "
The CLP value can be modified by the network, if a user's cell rate exceeds the negotiated cell rate. (For this purpose, the network executes a control function on the cell rate of every connection.) Changing the CLP =0 to CLP =1, is called cell tagging.
47 / 106
4.6
HEC
The HEC field has a double function :
" "
4.6.1
Error Multiplication
Why ? In the payload,1 bit error remains 1 bit error when the cell is switched further. It does not lead to more bit errors. On the other hand, 1 bit error in the header field can lead to multiple bit errors, when the cell is routed incorrectly. More precisely, a bit error in the VP/VC field can lead to cell loss and cell misinsertion. Such a case of a 1 bit error, finally leading to many bit errors is called error multiplication. Therefore, the header is protected with a Cyclic Redundancy Check (CRC), which is transmitted in the HEC field. This HEC field is checked at every ATM switch. The sending side calculates the HEC value : the polynomial generated by the first four bytes of the header is multiplied by 8 and then divided by x8 + x2 +x +1. The remainder of this division is transmitted in the HEC parameter field. At the receiving side, this HEC value is recalculated to check for errors. The HEC code is capable
"
to correct a single bit error to detect multiple bit errors In the normal (default) mode :
D D
"
HEC Algorithm
If a single bit error is detected, the error is corrected, and the state at the receiver switches to detection mode state. If a multiple bit error is detected the cell is discarded and the state at the receiver switches to detection mode state.
48 / 106
"
All cells with a detected header error are discarded and the receiver remains in the detection state mode. If a cell with a correct header is detected the receiver switches to the correction mode state.
Multiple bit error detected (cell discarded)
No error detected
No error detected
Single bit error detected ( header corrected ) C : Correction Mode D : Detection Mode
Figure 26
The reason for this dual kind of operation is the following : Trying to correct too many errors, introduces even more errors.Statistics show that most errors are either single bit errors or either lengthy burst errors.Thus, in case of 2 consecutive errored cell headers, chances are that it is a lengthy burst error and trying to correct is pointless.
49 / 106
4.6.2
Cell delineation
Cell delineation enables the receiver to recover the cell boundaries. To obtain this cell delineation, the correlation between the first four bytes of the cell header and the HEC value is examined. Figure 27 shows the cell delineation algorithm.
"
In the Synch state, it is assumed that the system is synchronised to the cell boundaries.
D
The Synch state is left (loss of cell delineation), when a number ( )of consecutive cells with incorrect HEC calculation result is found.
"
In the Hunt state, the delineation process checks bit by bit for the correctness of the HEC for the assumed header fields.
D D
If the calculation yields a correct result the receiver switches to the Presync state. If the calculation yields an incorrect result, the receiver remains in this state.
"
In the Presync state it is assumed that a correct cell delineation was found.
D
However, the calculation is continued for a further number ( ) of consecutive cells. If all consecutive cell headers yield correct, the Synch state is entered. If an incorrect HEC is found the receiver returns to the Hunt state.
50 / 106
H
consecutive incorrect HEC correct HEC incorrect HEC
P S
Exercise
As an exercise, you can calculate the probability for an ATM connection to get out of sync, because of 7 accidental errored headers. Suppose the BER = 10-5 What is the average time for this happening on an 155 Mbit/sec link ? P (1 wrong bit) = 10-5 P (1 correct bit) = 1 - 10-5 P (1 correct header) = (1 - 10-5)40 P (1 wrong header) = 1 - (1 - 10-5 ) 40 P (7 wrong headers) = (1 - (1 - 10-5 ) 40 ) 7
51 / 106
52 / 106
53 / 106
5.1
The Model
The B-ISDN Protocol Reference Model (BISDN PRM) is depicted as a cube and consists of 3 planes :
" " "
The User Plane for the transfer of User Data The Control Plane for the transfer of Signalling Messages The Management Plane
Each plane has specific functions; consequently they are separated from each other. Management plane Control plane
Higher layers
User plane
Higher layers
ATM Layer
Physical Layer
Figure 28
5.2
The Layers
The user plane and the control plane are organised into three layers :
" " "
54 / 106
Table 5 Functions
ATM Layers Main Layer ATM Adaptation Lay er (AAL) ATM Layer Sublayer Convergence Sublay er (CS) SAR
Convergence Segmentation and Reassembly (SAR) Generic Flow Control Cell Header Generation and Extraction Cell VPI / VCI Translation Cell Multiplexing / Demultiplexing Cell Rate Decoupling Header Error Control Cell Delineation Transmission Frame Alignment Transmission Frame Generation/Recovery Bit Timing Physical Medium
Physical Layer
5.2.1
PM
The PM performs functions, which are dependent on a physical medium, such as bit alignment and the generation of waveforms suitable for the medium. It is responsible for the correct transmission and reception of bits on the appropriate physical medium. This medium can be
" " "
At the lowest level, the functions of the PM are medium dependent, but this sublayer also does bit timing reconstruction at the receiver side. Because of this, the transmitter side is responsible for the insertion of the required bit timing information and line coding.
55 / 106
TC
The TC sublayer is responsible for all functions related to cell transmission. Its working depends on the used transmission system which carries the ATM cells. Examples of transmission systems are Plesiochronous Digital Hierarchy (PDH) and Synchronous Digital Hierarchy (SDH). Many different types of physical layer can be used to transport ATM cells, as shown in figure 29.
ATM
SDH
PDH
ADSL
...
Further, the TC sublayer is also responsible for the generation of the Header Error Control (HEC) syndrome of each cell at the transmitter, and its verification at the receiver. This allows cell delineation and detection of cell header errors. Finally, the TC performs cell rate decoupling. This function is used to adapt the useful information rate to the available payload of the transmission system. This decoupling of rates can be established by means of insertion of empty cells.
5.2.2
Cell multiplexing/demultiplexing. Cells from individual connections are multiplexed into one cell stream on the physical layer at the transmitter side. Cells are demultiplexed to individual cell streams at the receiving side.
"
VPI/VCI translation. The ATM layer translates incoming VPI/VCIs into appropriate outgoing VPI/VCIs when a cell passes through an ATM switch or cross-connect.
56 / 106
"
Cell header generation/extraction. The ATM layer at the transmitter side adds a header to the cell payload field received from the ATM Adaptation Layer. At the receiver side the cell header is extracted, and removed. The cell payload field is passed on to the ATM Adaptation layer.
"
Generic Flow Control at the UNI. The ATM layer controls the traffic flow at the UNI to control overload conditions that can occur in the network.
5.2.3
The AAL
The AAL performs the adaptation of higher layer protocols to the fixed size ATM cells and it enhances the services provided by the ATM layer to the requirements of the higher layers. The AAL is service dependent and also foresees error detection and correction, time/clock recovery, etc.
AAL 1
AAL 2
AAL 3/4
AAL 5
...
ATM
Figure 30 Different AALs
AAL-5 supports data services. Its functionality is simplified to support high speed data transfer in an efficient way. It assumes that the layers above AAL perform error recovery, retransmission and sequence numbering. Consequently it does not provide this functionality. As a result, the overhead information is minimal. AAL-5 is also referred to as .Simple and Efficient Adaptation Layer (SEAL)
57 / 106
The SAR sublayer segments the information into 48 octets which will be transported inside ATM cells. The cell transporting the final part can be identified via the PTI3 bit =1. However, before this segmenting is done, some additional fields are added to the information
Information Padding
(0...47 octets)
Trailer
(8 octets)
CPI
0000 0000
Length
(of the information)
CRC
Figure 31
The AAL-5
58 / 106
5.3
User Plane
As far as the user plane is concerned, ATM switches are not interested in the ATM cell payload, i.e. the information is transported transparently Thus, for ATM switches and for user information, the AAL is not important. Of course, it is important for the end-user equipment.
Control Plane
However, for the control plane it is different. This control information (signalling messages) has to be examined by every ATM switch. The AAL used, also called the Signalling ATM Adaptation Layer (SAAL) is AAL5. ATM cells belonging to the user plane and the control plane have to be discerned. An ATM cell of the control plane, carrying signalling information, can be distinguished because VCI = 5. Control Plane
Higher layers
User Plane
SAAL
VCI = 5 VCI > 32
Figure 32
59 / 106
60 / 106
6 Call Configurations
6 Call Configurations
This chapter describes 2 call configurations, specific for ATM, i.e. : G G Multi-connection Calls Point-to-Multipoint Connections
61 / 106
6 Call Configurations
6.1
Multi-Connection Calls
In B-ISDN, a clear separation exists between the concepts :
" "
Call Connection
A call is setup by a signalling dialogue. However, signalling messages are designed in a way that multiple connections can be requested simultaneously. These multiple connections could serve to transport multiple service-components of a call, such as :
" " " "
For this purpose, the used identification of virtual connections is specifically useful, more precisely : Different service components can be multiplexed on the same VP .
62 / 106
6 Call Configurations
6.2
Point-to-Multipoint Calls
Figure 33 shows an example of a point-to-multipoint call, in which a number of receivers (the TV-sets) receive the images from a sender (the camera). Note that all leafs will receive the same image at any time. The leafs can not send information back to the sender, nor can they send information to each other. Thus, such a point-to-multipoint connection is always unidirectional. The source of a point-to-multipoint is called the "root" of the connection, while the destination sides are called the "leafs".
Leaf 1
Leaf 2
Root
Leaf 3 Leaf 4
Figure 33
Setup
The point-to-multipoint connection is setup by first requesting the establishment of a connection between the root and one leaf. After this setup has progressed, additional leaves can be added to the connection by add party requests from the root. These parties can be added "in parallel", i.e the root does not need to wait for a response related to one add party request before issuing the next one. Exchanges providing point-to-multipoint connections need a replication function. For efficiency, the replication should be performed as close to the leafs as possible.
Replication
63 / 106
6 Call Configurations
64 / 106
65 / 106
7.1
Connection Admission Control (CAC) Usage Parameter Control (UPC) / Network Parameter Control (NPC). This functionality is often referred to as Policing.
7.1.1
CAC
Use
CAC is defined as the set of actions taken by the network at the call establishment phase in order to establish whether a connection can be accepted or rejected. Moreover, if a new connection is accepted, the Quality of Service (QoS) of existing connections should be maintained. Inputs and outputs are shown in figure 34. Inputs
Existing Connections Connection Characteris tics Buffer Sizes Output Link Rate
Outputs
CAC
Yes No
Figure 34
CAC functionality
The used CAC algorithm is not standardized Therefore, it will be an area of competition for the operators. Indeed, a better CAC can represent more money, because of a better utilization of bandwidth.
66 / 106
Traffic Descriptors
The basis for the CAC must be some kind of description of the traffic the user is intending to send. This description will use traffic parameters. Examples are :
" " " "
Peak Cell Rate (PCR) Sustainable Cell Rate (SCR) Intrinsic Burst Tolerance (IBT) Minimum Cell Rate (MCR)
PCR
SCR
Figure 35
Traffic Parameters
More traffic parameters may be defined in the future. However, it is important that every parameter is :
"
Simple and understandable in an unambiguous way to the user terminal and to the network. Useful for CAC algorithms Enforceable by the UPC / NPC.
" "
67 / 106
A set of traffic control functions, based upon a well-defined set of traffic parameters is called an ATM Transfer Capability (ATC). From the operator's perspective, an ATC may provide gain through statistical multiplexing. On the other hand, a user can have a preference for an ATC because of pricing.
DBR
The basic (and default) ATC is the Deterministic Bitrate (DBR). If a user requests a DBR connection, he needs to specify (via signalling procedures) :
"
PCR
DBR is intented to be used to meet the requirements of constant bitrate traffic.. It provides a static amount of bandwidth that is continuously available during the connection lifetime. Thus the source can emit cells at the PCR at any time and for any durations and the QoS commitments still pertain. It provides QoS with respect to CLR and Cell Transfer Delay (CDT).
SBR
A second ATC is .Statistical Bitrate (SBR). If a user requests an SBR connection, he needs to specify (via signalling procedures) :
" " "
SBR is intented to be used to meet the requirements of variable bitrate. It can be used if the source can describe in greater detail than just the PCR, the cell flow that it will emit. SBR provides QoS with respect to CLR and optionally to CDT.
Other ATC's
Available Bitrate (ABR) ATM Block Transfer (ABT) Unspecified Bitrate (UBR) Guaranteed Framerate (GFR)
68 / 106
7.1.2
UPC / NPC
Use
UPC / NPC is the set of actions taken by the network to monitor and control that the traffic contract is respected in terms of traffic offered and validity of the ATM connection, at the user acces and the network access, respectively. The main purpose is to protect network resources from malicious as well as unintentional misbehaviour wich can affect the QoS of other connections. The policing function should be located as close as possible to the source of the traffic, because detection of excessive traffick must be done as early as possible. However, it should not be located in the subscriber terminal, because of the user can not be trusted. The location of UPC / NPC is shown in figure 36.
Where
UPC
NPC
Figure 36
Actions
Conforming Non-conforming
This happens in a standardized way. The policing function takes no actions on conforming cells. On the other hand, actions upon non-conforming cells (which are operator specific ) may be :
" " "
Compliant Not-compliant
69 / 106
When a connection contains too many non-conforming cell, she is determined to be non-compliant. The point at which this happens is also not standardized. For a non-compliant connection UPC / NPC may initiate connection release.
The ideal UPC / NPC function (and discarding non-conforming cells) will only admit the contracted PCR. Thus it would look as in figure 37.
Output Traffic
PCR
Input Traffic
Figure 37
The GCRA
To classify each cell as conforming or not, a real time standardized algorithm has to be performed, with as input some statistical (long-term) description of the traffic. This standardized algorithm is called the Generic Cell Rate Algorithm (GCRA). It classifies the cells as shown in figure 38, using the inter-arrival times.
Last Conforming Cell
TAT -
TAT
Non - Conforming
Conforming
Figure 38
Starting from the last conforming cell, the expected time of the next cell can be calculated. This is the Theoretical Arrival Time (TAT). For instance, for a DBR connection, the expected interval T between cells would be : T = 1 / PCR
70 / 106
Cells arriving too early" can then be classified as non-conforming. However, a certain tolerance is granted, i.e. cell arriving before TAT but still within the interval [TAT - , TAT] are still considered to be conforming. is called the Cell Delay Variation Tolerance (CDVT). Cells, sent in a conforming way, may exhibit some Cell Delay Variation (CDV) because of a number of reasons,. like
" " "
Non-conforming cell
Yes
ta < TAT -
No
Conforming cell
For SBR connections, things become more complicated. There we need to run 2 GCRA algorithms in parallel : GCRA ( 1/PCR, ) AND GCRA ( 1/SCR, IBT)
71 / 106
7.1.3
ABR
Use
ABR is a transfer capability, intended to support sources which have the ability to adapt their transfer rate. Thus, ABR is intended for the transfer of data. An ABR source is allowed to send at a certain bitrate, consisting of
" "
QoS
no garanty upon cell delay. a very strong garanty upon cell loss.
Feedback Loop
The size of the variable bandwidth has to be conveyed to the ABR source, which has to obey. For this purpose, a double feed-back loop is setup, consisting of Resource Management (RM) cells. These RM cells can be considered to be transport vehicles for the ATM switches to communicate their feedback information.
" " " "
RM cells are generated by the source RM cells are looped back by the destination RM cells are in-band. RM cells are in the forward or backward direction.
= =
Figure 40
End-to-end
72 / 106
"
Link-by-link
In the case of a link-by-link feedback loop, ATM exchanges have to behave as a Virtual Source (VS) and a Virtual Destination (VD). The use of VS/VD has the following advantage
"
An operator can have a full control of the loop in his own network
Backward : Faster Feedback Forward : Other switches can use the feedback.
The feedback an ATM switch writes into an RM cell can be, either :
" "
Binary Explicit
Binary feedback is indicated by the values of only 2 bits, being Congestion Indication (CI) and No Increase (NI). The following are the rules for binary feedback. Table 6 CI 0 0 1 1 NI 0 1 0 1 ABR Binary Feedback Action ACR = ACRold + AIF * PCR ACR = ACRold ACR = ACRold - RDF * ACRold
73 / 106
Because of these rules, traffic increase is slow but a traffic decrease is fast. The advantage of this is better network protection Feedback, written in RM cells may not overrule the feedback of a previous node, unless the new feedback is more demanding, i.e. restricts the source to an even lower bandwidth. This is shown in figure 41.
90 20 60
90 20 20
90 50 80
90 50 50
Figure 41
Policing
Also the traffic of ABR sources has to be policed, because now the allowed traffic rate is variable, also the parameters, to be used by the GCRA will be variable. We call this new algorithm the DGCRA = Dynamic GCRA It takes into account the minimum (2) and maximum delay (3) for the feedback to take effect.
Fairness
In distributing traffic over all ABR connections, ATM switches have to stick to the concept of fairness, i.e. no particular source may be privileged. Fairness algorithms are not standardized. Several possibilities are :
" " "
Total Bandwidth / N Total Bandwidth * MCRi / MCRi Total Bandwidth * ( PCRi - MCRi ) / (PCRi -MCRi)
74 / 106
8 B-ISDN Signalling
8 B-ISDN Signalling
This chapter describes very briefly the signalling protocols and their stacks used in a B-ISDN
75 / 106
8 B-ISDN Signalling
8.1
Classification
Topological Classification
Regarding a network topology, 3 signalling interfaces exist :
" " "
UNI = User-to-Network interface NNI = Network-to-Network interface ANI = Access Network to Network interface
UNI
Access Network
ANI
Figure 42
Signalling Interfaces
The following signalling systems are used at these interfaces. Table 7 UNI DSS1 DSS2 Signalling Systems NNI ISUP (CCS #7) B-ISUP (CCS #7) ANI V5 VB5
In STM, a user can not, but use only his allocated piece of bandwidth, i.e., the channel(s) allocated to him. This channel is agreed upon at connection setup, and is communicated by means of the existing N-ISDN protocols. Also in B-ISDN, an agreement must be made between the user and the network. This agreement is called the User-Network Traffic Contract". Unlike N-ISDN, much more traffic aspects must be agreed upon. The amount, speed, burstiness and other traffic aspects of the user information (which will be transmitted) must be communicated by the B-ISDN protocols. This aspect increases the required capability of the B-ISDN signalling procedures. Indeed, a traffic contract must be
76 / 106
8 B-ISDN Signalling
communicated or even negotiated. The following are possible parameters describing the characteristics of an ATM connection :
" " " " " "
ATM Traffic Descriptor OAM Traffic Descriptor Quality of Service Parameters Type of AAL Communication Configuration etc.
77 / 106
8 B-ISDN Signalling
8.2
8.2.1
Protocol Stacks
DSS2
Above the ATM layer, which provides transport of ATM cells without any processing like error control, there is an Signalling ATM Adaptation Layer (SAAL). This SAAL layer is further divided into :
" "
an ATM Adaptation Layer 5 (AAL-5) Common Part the Service Specific Connection Oriented Protocol (SSCOP) layer the Service Specific Coordination Function (SSCF) layer
"
The highest layer in DSS2 is described in Q.2931, which is the BB version of NB Q.931, also called DSS2 Layer 3. It describes all messages and procedures for point-to-point connections. It can be extended by Q.2961 (for additional parameters and procedures) and by Q.2971 (which describes the procedures for point-to-multipoint call control). The protocol stack for DSS2 is shown in figure 43. Q.2931/61/71. Q.2130 Q.2110 I.363.5 I.361 I.432 Figure 43
SSCF SSCOP AAL5 ATM Layer Physical Layer
Error Detection
78 / 106
8 B-ISDN Signalling
8.2.2
B-ISUP
At the NNI, either the ATM network or, as a national option, the existing CCS #7 network can be used for signalling information transfer.
Option 1
Messages and Parameters Addressing and Network Management Filter Error Correction Error Detection
SAAL
Note that a lot of layers are the same as the ones for the DSS1 stack. Only the 3 highest layers differ.
Option 2
However, as an intermediate solution, an operator can decide to use his existing CCS #7 network, to provide for the B-ISDN signalling. This means he will reuse his narrowband links, which follow the classical CCS #7 protocol stack. Only B-ISUP will now replace the narrowband ISUP The stack for this option is shown in . figure 45.
79 / 106
8 B-ISDN Signalling
Q.2761-Q.2764
Messages and Parameters Addressing and Network Management Error Detection and Correction
80 / 106
81 / 106
9.1
Both are digital transmission systems, multiplexing multiple information streams into higher order hierarchies. The difference between them lies in how the different network nodes are synchronized.
Plesiochronous Network Elements use as a clock :
PDH
Synchronous Transmission
SDH
Figure 46
PDH
In PDH, every network uses its own clock, which will never be completely the same as any other clock. Thus, network elements work almost synchronous or Plesiochronous". Therefore, when multiplexing different lower order streams together, stuffing bits" have to be inserted. These stuffing bits provide the flexibility needed. They can contain information from a source, or they can be empty, i.e. not containing any user information. In this way, small clock differences ("50 ppm) can be handled. The basic European frame (called an E1) contains 32 channels of 64 Kbit/sec. It is shown in figure 47.
82 / 106
15 16 17
31 0
Synchronization
Signalling Information
1 frame = 125 sec 8 bits = 64 Kbit/sec 125 sec 32 * 64 Kbit/sec = 2,048 Mbit/sec
PCM Structure
The European way is to multiplex 4 lower order bitstreams into 1 higher order bitstream. However, because of historical reasons, a different PDH hierarchy was defined in North-American (and yet another different one in Japan). The European and the North-American multiplexing hierarchies are shown in table 8. Table 8 PDH Multiplexing Hierarchies European First Order Third Order Fourth Order Note E1 E3 E4 2, 048 8, 448
Mbit/s Mbit/s Mbit/s Mbit/s
Second Order E2
34, 368
44, 736
138, 264
139, 264
The bitrate of the higher order is always more than n times the lower bitrate. This is because some overhead is introduced, like stuffing bits and frame alignment patterns A big disadvantage of PDH is the impossiblity to reach, for example, a particular E1 out of an E4. The inverse demultiplexing way has to be performed completely for this. This is called back-to-back multiplexing.
83 / 106
140 Mbit/s
140 Mbit/s
34 Mbit/s
34 Mbit/s
34 Mbit/ s 34 Mbit/ s
34 Mbit/s
64 x 2 Mbit/s
Figure 48
Back to Back Multiplexing A lot of equipment is needed for this procedure, a cascade of multiplexers/demultiplexers has to be installed in the network node. Therefore, PDH has a number of disadvantages :
"
expensive, because tailor-made installation is required in each network node. lots of processing increases the probability of faults and failures.
"
SDH
To overcome these problems, in the SDH network, the network elements are fully synchronised to a master clock. This permits a more appropriate way to remove and add lower order signals from and into a higher order signal. It is now possible to directly drop/add lower order signals, without having to demultiplex/multiplex through the whole hierarchy. This is called add / drop multiplexing.
84 / 106
ADD/DROP multiplexer
Drop
e.g. : 2 Mbit/s
Add
Figure 49
Add/Drop Multiplexing
Because of this feature, SDH networks will typically be ring shaped. The ring carries information destined for different nodes. This lower order information is removed form the high order signal in that specific Add/Drop Multiplexer (ADM).
140 Mbit/s ADM
ADM
ADM
ADM
Figure 50
Further, for the new SDH network, only 1 common higher order signal is defined, which is used worldwide. This new higher order signal is compatible with the existing PDH tributaries. This means, that it has to be able to carry lower order signals of the different PDH systems. This is shown in table 9.
85 / 106
Table 9
SDH Containers Bitrate 1544 Mbps 2048 Mbps 8 Mbps 34 Mbps 140 Mbps
The existing PDHs only define bit rates up to 140 Mbit/s. Modern optical systems can offer higher bit rates. So, in the SDH new levels of multiplexing hierarchies are defined for these high bit rates. This is shown in table 10. Table 10 The SDH Hierarchy Bitrate 155.52 Mbps 622.08 Mbps 2.488 Gbps 10 Gbps 40 Gbps
The bitrate of the higher order is always exactly 4 times the lower bitrate. This is because no overhead is introduced. To conclude, we summarize the properties of the PDH and SDH systems in figure 51 and table 11.
86 / 106
PDH
SDH
Stuffing Bits
Figure 51 Table 11
PDH versus SDH PDH versus SDH PDH SDH Synchronous Flexible Hierarchy Very limited bit-stuffing Pointers to frame-boundaries Add-Drop multiplexing One Higher order Hierarchy Extended management features
Plesiochronous Fixed Hierarchy Stuffing Bits No pointers to frame-bound aries Back-to-back multiplexers Different Hierarchies Limited management features
87 / 106
9.2
9.2.1
Header Header
Figure 52
88 / 106
9.2.2
59 bytes
Figure 53
FA1 / FA2 : Frame Alignment Signal EM : Error Monitoring. TR : Trail Trace. It identifies the trail between transmitter and receiver. MA : Maintenance and Adaptation byte.
D D D D D
"
bit 1: FERF bit 2: FEBE bit 3 to 5 : payload type (010 for ATM cells) bit 6 to 7 : payload dependent (for example: TU multiframe indicator) bit 8: timing marker, refers to the timing source primary reference clock relationship
" "
Figure 54 shows the ATM cells in a 34 368 kbit/s frame. The ATM cells are mapped into the 530 payload bytes of the frame.
89 / 106
Header Header Header Header 9 bytes Header Header Header Header 59 bytes Header Header Header
Figure 54
90 / 106
9.3
270
Overhead
PAYLOAD
9 rows
125 sec
Figure 55
STM-1
Content : 9 x 270 bytes = 2430 bytes Period : 125 sec; Bitrate : 155,520 Mbit/sec 2430 * 8 bits in every 125 sec Payload capacity : 150,336 Mbit/sec 2349 * 8 bits in every 125 sec
"
The payload area of the STM-1 contains Virtual Containesr (VC) of different sizes : VC11, VC12, VC2, VC3, VC4. ATM cells are mapped into these VCs. Any type of VC can carry ATM cells. The mapping is performed in such a way, that the ATM cell's byte boundaries are aligned with the VCs byte boundaries Idle cells are inserted if the offered rate of the source is not sufficient to fully load the SDH capacity. Thus, the actual transmitted cellstream has a rate that is synchronous with the SDH Container in which it is transported, although the information rate is defined by the ATM source. To prevent the ATM cell payload to accidentally contain the SDH frame alignment or ATM cell delineation information stream, the
91 / 106
payload part of the ATM cell is scrambled. This also protects the SDH network and the users against malicious users Because a container's capacity is not an integer multiple of the cell size, a cell can go across a container's boundary. To find the ATM cell boundary (cell delineation) the Header Error Control (HEC) parameter is used, which is part of the contents of the ATM cell header. Mapping of ATM cells is done mainly into a VC-4. Figure 56 shows an example.
POH
J1 B3 C2 G1 F2 H4 Z3 Z4 Z5 260 261 ATM cell
VC-4
Figure 56
In the Path Overhead (POH) of the VC-4, 2 bytes are important, for the mapping of ATM cells.
"
C2 : indicates that the VC-4 contains ATM cells H4 : pointer to the beginning of the first cells
"
Besides the H4 byte, also another cell delineation mechanism may be used , the HEC algorithm A VC-4 has a payload with a size of 9 x 260 = 2340 bytes. Thus, it can contain 44 complete ATM cells. The last cell will not be complete, only 8 bytes are in one payload, while it continues in the next VC-4.
92 / 106
9.4
IMA
Why IMA
Often ATM's biggest challenge is cited to be competition in the Local Area Network (LAN) area, coming from fast Ethernet and Gigabit Ethernet. But ATM faces a problem on the Wide Area Network (WAN) as well. Not at the core of public networks, but at their edges. Only few companies can afford E3, T3 or STM-1 pipes to link their offices. Inverse multiplexing lets companies who can not afford large and expensive lines, to treat several smaller lines as one single logical connection. The fact that a number of smaller lines are used should be transparent to the application and to the rest of the network, cell order and format is retained. An inverse multiplexer takes traffic from a relatively high-bandwidth connection (say, a campus ATM backbone running at 155 Mbit/sec) and spreads it accross multiple lower speed WAN circuits. This offers several advantages :
"
Rather than paying for 45 or 155 Mbit/sec WAN capacity, net managers can now deploy exactly as much WAN bandwidth as they need, by gradually employing more E1 or T1s. This allows for a cost-effective migration. Maybe even more important, applications can use the bandwidth more effectively, since idle capacity on one line can be combined with unused capacity on another. For example, 5 connections of 1,2 Mbit/sec.
D D
"
Without IMA, each connection has to be mapped to its own T1, for a total of 5 T1 links. With IMA, the total needed capacity of 6 Mbit/sec can be spread over only 4 T1 links.
IMA spread the ATM traffic over multiple links, called an IAM group. In the transmit direction, the cell stream is distributed on a cell basis, across the multiple links. At the receiving side, the receiving IMA recombines the cells from each link, recreating the original ATM cell stream. The ATM cells are spread in a cyclical fashion over the links (i.e. cyclic round-robin). This is shown in figure 57.
93 / 106
Figure 57
IMA Objectives
Provide for the provisioning of bandwidth in physical link increments (e.g. E1 increments). Provide transparent transport of the ATM layer and higher layer signals (including the preservation of cell order, control of cell delay variation, etc..) Use a cell based multiplexing technique Detect and reject lines with greater delay than the provisioned maximum differential delay tolerance. Add/delete links while maintaining the IMA group in service. Support of all ATCs.
"
" "
" "
IMA Group
The IMA interface, called an IMA group, consists out of a number of N transmission links, each operating at the same link cell rate. Each link in the IMA group is identified by a unique Link Identifier (LID) between 0 and 31. The ATM cells shall be distributed over these N links, in cyclic round robin fashion, using an ascending order based on the assigned LIDs. The unit of control in IMA is called an IMA frame. An IMA frame consists out of a number M consecutive cells, on each link. The transmission of IMA frames must be aligned on all links within an IMA group, as is shown in figure 58. Every IMA frame contains a newly defined type of OAM cell, called an IMA Control Protocol (ICP) cell. An ICP cell indicates its relative offset position with respect to the beginning of an IMA frame. Differential link delay can cause the reception to be misaligned in time. The receiving IMA equipment can recover this IMA frame alignment by means of these ICP cells. Thus, cell sequence and cell delay can be preserved.
IMA Frame
94 / 106
Figure 58
IMA Frames
95 / 106
96 / 106
Abbreviations
Abbreviations
AAL AAL-5 ABR ABT ADM ATC ATM B-ISDN BER CAC CDT CDV CDVT CI CL CLP CLR CO CRC CS DBR GCRA GFC GFR HDTV HEC IBCN IBT ICP IMA LAN ATM Adaptation Layer ATM Adaptation Layer 5 Available Bitrate ATM Block Transfer Add/Drop Multiplexer ATM Transfer Capability Asynchronous Transfer Mode Broadband Integrated Services Digital Network Bit Error Rate Connection Admission Control Cell Transfer Delay Cell Delay Variation Cell Delay Variation Tolerance Congestion Indication Connectionless Cell Loss Priority Cell Loss Ratio Connection Oriented Cyclic Redundancy Check Convergence Sublayer Deterministic Bitrate Generic Cell Rate Algorithm Generic Flow Control Guaranteed Framerate High Definition Television Header Error Control Integrated Broadband Communication Network Intrinsic Burst Tolerance IMA Control Protocol Inverse Multiplexing for ATM Local Area Network
97 / 106
Abbreviations
LID MCR NI NNI NPC PCR PDH PM POH PSTN PTI QoS RM SAAL SAR SBR SCR SDH SEAL SSCF SSCOP STM STM TAT TC TDM UBR UNI UPC VC VCI VD VOD
Link Identifier Minimum Cell Rate No Increase Network-to-Network Interface Network Parameter Control Peak Cell Rate Plesiochronous Digital Hierarchy Physical Medium Path Overhead Plain Old Telephone Service Payload Type Identifier Quality of Service Resource Management Signalling ATM Adaptation Layer Segmentation and Reassembly Statistical Bitrate Sustainable Cell Rate Synchronous Digital Hierarchy Simple and Efficient Adaptation Layer Service Specific Coordination Function Service Specific Connection Oriented Protocol Synchronous Transfer Mode Synchronous Transport Module Theoretical Arrival Time Transmission Convergence Time Division Multiplexing Unspecified Bitrate User-to-Network Interface Usage Parameter Control Virtual Containesr Virtual Cell Identifier Virtual Destination Video On Demand
98 / 106
Abbreviations
Video on Demand Virtual Path Identifier Virtual Source Wide Area Network
99 / 106
Abbreviations
100 / 106
Glossary
Glossary
101 / 106
Glossary
102 / 106
Appendix A References
Appendix A
References
1. I.113 :Vocabulary of Terms for Broadband Aspects of ISDN. 2. I.121 :Broadband Aspects of ISDN 3. I.432 : B-ISDN user-network interface - physical layer specification 4. I.150 : B-ISDN ATM Functional Characteristics 5. I.211 : B-ISDN Service Aspects. 6. I.321 : B-ISDN Protocol Reference Model and Applications. 7. I.327 : B-ISDN Network Functional Architecture. 8. I.361 : B-ISDN ATM Layer Specification. 9. I.362 : B-ISDN ATM Adaptation Layer (AAL) Functional Description. 10. I.363 : B-ISDN ATM Adaptation Layer (AAL) Specification. 11. I.371 : Traffic and Congestion Control in B-ISDN. 12. G.701 : Vocabulary of digital transmission and multiplexing, and PCM terms. 13. G.702 : Digital hierarchy bit rates. 14. G.703 :Physical/electrical characteristics of hierarchical digital interfaces. 15. G.704 : Synchronous frame structures used at primary and secondary hierarchical levels. 16. G.732 : Characteristics of primary PCM multiplex equipment operating at 2 048 kbit/s. 17. G.742 : Second order digital multiplex equipment operating at 8 448 kbit/s and using positive justification 18. G.751 : Digital multiplex equipments operating at the third order bitrate of 34 368 kbit/s and the fourth order bitrate of 139 264 kbit/ and using positive justification. 19. G.804 : ATM cell mapping into plesiochronous digital hierarchy (PDH). 20. G.832 : Transport of SDH elements on PDH networks: frame and multiplexing structures. 21. G.707 : Synchronous digital hierarchy bit rates. 22. G.708 : Network node interface for the synchronous digital hierarchy. 23. G.709 : Synchronous multiplexing structure.
103 / 106
Appendix A
References
104 / 106
Fax E Mail
+49/711/821-4 75 23 J.Schwarz@alcatel.de
Important!
This is not a fault report! In case of documentation errors, the officially agreed fault reporting mechanism has to be used, in addition.
Document Title
Document Identification
1.
Please rate the effectiveness of hi document i the f ll i this d in h following areas. areas Please mark ` ' Pl k `x': ( 1 ) - Excellent ll ( 2 ) - Good ( 3 ) - Fair ( 4 ) - Poor ( 5 ) - Not applicable
105 / 106
Document Identification
2.
In your opinion, how could we improve this document? Please mark `x' and provide details for the suggested improvement . [ ] [ ] [ ] [ ] [ ] [ ] Add more illustrations Add more examples Add more detail Make it shorter Make it less technical Improve ...
Details/Comments: (Use separate sheet, if necessary.)
3.
What is this document's best feature? Feel free to write any additional comments as well.
(Use separate sheet, if necessary.)
4.
If we may contact you to clarify your comments, please complete the following: Name Company Address Phone E Mail Fax Date
5.
Please return your comments to Alcatel TELECOM. (For address see first page of this questionnaire.) Thanks in advance for your assistance.
106 / 106