Transcript Document

Chapter 9
High-Speed
Digital Access:
DSL, Cable Modem,
and SONET
McGraw-Hill
1
©The McGraw-Hill Companies, Inc., 2004


Traditional modems impose an upper limit on the data rate available.
DSL Technology


DSL provide higher-speed access to the Internet over the existing local
loops. Set of technologies: ADSL, VDSL, HDSL and SDSL.
ADSL: Asymmetric DSL







McGraw-Hill
Provides higher speed (bit rate) in the downstream direction (from the Internet
to the resident) than in the upstream direction (from the resident to the
Internet). So, it is called as asymmetric.
Designed for residential users; it is not suitable for businesses.
ADSL provides data rate more than traditional modems.
The existing local loops can handle bandwidths up to 1.1 MHz but the filter
installed at the end of the line by the telephone company limits the bandwidth
to 4 KHz (sufficient for voice communication). This was done to allow the
multiplexing of a large number of voice channels.
Unfortunately, 1.1 MHz is just the theoretical bandwidth of the local loop.
Factors such as the distance between the residence and the switching office,
the size of the cable, the signaling used, and so on affect the bandwidth.
Designers are aware of this problem and used an adaptive technology that
tests the condition and bandwidth availability of the line before setting on a
data rate.
The data rate of ADSL is not fixed; it changes based on the condition and type
of the local loop cable.
2
©The McGraw-Hill Companies, Inc., 2004
Discrete Multitone Technique- DMT







McGraw-Hill
Modulation technique of ADSL that combines QAM and FDM.
Each system can decide on its bandwidth division.
Voice: Channel 0 is reserved for voice communication.
Idle: Channel 1-5 are not used to allow a gap between voice and data
communication.
Upstream data and control: Channel 6 - 30 (25). One channel for
control and 24 for upstream data, each using 4KHz.
 Upstream bandwidth = 24*4000*15 = 1.44Mbps
Downstream data and control: Channel 31 – 255 (225). One channel
for control and rest are for downstream data.
 Downstream bandwidth = 224*4000*15 = 13.4 Mbps
Actual bit rates are much lower than the above-mentioned rate:
 Upstream = 64 Kbps to 1 Mbps
 Downstream = 500 Kbps to 8 Mbps
3
©The McGraw-Hill Companies, Inc., 2004
ADSL


McGraw-Hill
Customer site: ADSL Modem

Local loop connects to the filter which separates voice and data
communication.

ADSL modem modulates the data, using DMT, and creates
downstream and upstream channels.
Telephone company site: DSLAM

Instead of an ADSL modem, a device called a digital subscriber line
access multiplier (DSLAM) is installed that functions similar to ADSL

DSLAM packetizes the data to be sent to the Internet (ISP Server)
4
©The McGraw-Hill Companies, Inc., 2004
Other DSL Technologies

SDSL – Symmetric DSL



HDSL – High-Bit-Rate DSL






Suitable for businesses.
It divides the available bandwidth equally between down and up streams.
Alternative to T-1 Lines (1.544 Mbps).
T-1lines use AMI (Alternate mark inversion) encoding which is very
susceptible to attenuation at high frequencies. This limits the length of a T1 line to 1 km. For long distances, we need repeaters.
Uses 2B1Q encoding, which is less susceptible to noise.
Provides data rate of almost 2 Mbps and upto 3.6 km without repeater
Use two twisted-pair wires for full-duplex transmission
VDSL - Very-high-bit-rate DSL


McGraw-Hill
Similar to ADSL. Use coaxial, fiber-optic, or twisted-pair cable for short
distance (300 to 1800m).
Modulation technique used is DMT with bit rate of 50 to 55 Mbps
downstream and 1.5 to 2.5 Mbps upstream.
5
©The McGraw-Hill Companies, Inc., 2004
Cable Modem


McGraw-Hill
DSL provides high data rate using existing unshielded twisted-pair cable,
which is very susceptible to interference. This imposes an upper limit on
the data rate.
Traditional Cable Networks

Called as Community Antenna TV (CATV) because an antenna at the top of a
tall hill or building received the signals from the TV station and distributed
them, via coaxial cables, to the community.

Cable TV office called as head end, receives video signals from broadcasting
stations and feeds the signals into coaxial cables.

As the signals became weaker, amplifiers are installed. There can be up to 35
amplifiers between head end and the subscriber premises.

At other end, splitters split the cable, and taps and drop cables make the
connections to the subscriber premises.

Due to attenuation of signals and use of a large number of amplifiers,
communication in the traditional cable TV network is unidirectional
6
©The McGraw-Hill Companies, Inc., 2004
Hybrid Fiber-Coaxial (HFC) Network







McGraw-Hill
Network uses a combination of fiber-optic and coaxial cable.
Transmission medium from the cable TV office to a box, called the fiber
node, is optical fiber; from the fiber node through the neighbourhood and
into the house is still coaxial cable.
Regional Cable Head (RCH) normally serves up to 400,000 subscribers.
RCHs feed the distribution hubs, each of which serves up to 40,000
subscribers.
Modulation and demodulation of signals are done at the distribution hubs;
The signals are then fed to the fiber nodes through fiber-optic cables.
Fiber node splits the analog signals so that the same signal is sent to each
coaxial cable. Each coaxial cable servers up to 1000 subscribers.
The use of fiber-optic cable is to reduce the need for amplifiers down to
eight or less.
Communication in an HFC Cable TV network can be bidirectional
7
©The McGraw-Hill Companies, Inc., 2004
Coaxial Cable Bands in HFC







McGraw-Hill
Coaxial cable has a bandwidth that ranges from 5 to 750 MHz (approx).
Cable company has divided this bandwidth into three bands: video,
downstream data, and upstream data.
The downstream-only video band occupies frequencies from 54 to 550 MHz.
Since each TV channel occupies 6 MHz, this can accommodate more than
80 channels.
The downstream data (from the Internet to the subscriber premises)
occupies the upper band, from 550 to 750 MHz. This band is also divided
into 6-MHz channels.
Downstream data are modulated using the 64-QAM (or possibly 256-QAM)
modulation technique.
Upstream data (from the subscriber premises to the Internet) occupies the
lower band, from 5 to 42 MHz. This band is also divided into 6-MHz
channels.
Upstream data band uses lower frequencies that are more susceptible to
noise and interference. Thus QAM is not suitable and so QPSK is used.
8
©The McGraw-Hill Companies, Inc., 2004
Coaxial Cable Bands in HFC

Data Rate for downstream:
6 bits for each baud in 64-QAM.

One bit is used for forward error correction; this leaves 5 bits of data per baud.

The standard specifies 1 Hz for each baud; this means that, theoretically,
downstream data can be received at 30 Mbps (5 bits/Hz * 6 MHz). The standard
specifies only 27 Mbps. Since the cable modem is connected to computer
through a 10base-T cable, this limits the data rate to 10 Mbps.
Data Rate for Upstream:

2 bits for each baud in QPSK.

The standard specifies 1 Hz for each baud; this means that, theoretically,
downstream data can be received at 12 Mbps (2 bits/Hz * 6 MHz). However, the
data rate is usually less than 12 Mbps.


McGraw-Hill
9
©The McGraw-Hill Companies, Inc., 2004
Coaxial Cable Bands in HFC

McGraw-Hill
Upstream and Downstream sharing by subscribers:

Upstream Sharing:
 Upstream data bandwidth is 37 MHz. Only six 6-MHz channels are
available.
 A subscriber needs to use one channel to send data in upstream direction.
 To share among many subscribers, we use time sharing. The band is divided
into channels using FDM; these channels must be shared between
subscribers in the same neighbourhood. The cable provider allocates one
channel, statically or dynamically, for a group of subscribers. Subscribers
content for a channel to send data.

Downstream Sharing:
 Downstream band has 33 channels of 6 MHz.
 A cable provider probably has more than 33 subscribers; therefore each
channel must be shared between a group of subscribers.
 We have a multicasting situation here: If there are data for any of the
subscribers in the group, the data are sent to that channel. Each subscriber is
sent the data. But since each subscriber also has an address registered with
the provider, the cable modem for the group matches the address carried
with the data to the address assigned by the provider. If the address
matches, the data is kept; otherwise, they are discarded.
10
©The McGraw-Hill Companies, Inc., 2004
CM and CMTS



McGraw-Hill
To use a cable network for data transmission, we need two key devices: a
CM and a CMTS.
Cable Modem – CM

Installed on the subscriber premises and similar to an ADSL modem.
Cable Modem Transmission System – CMTS

Installed inside the distribution hub by the cable company.

It receives data from the Internet and passes them to the combiner,
which sends them to the subscriber. The CMTS also receives data from
the subscriber and passes them to the Internet.
11
©The McGraw-Hill Companies, Inc., 2004
Data Transmission Schemes: DOCSIS



Data over Cable System Interface Specification (DOCSIS) designed by
Multimedia Cable Network Systems (MCNS).
DOCSIS defines all the protocols necessary to transport data from a CMTS to
a CM.
Upstream communication








CM checks the downstream channels for a specific packet periodically sent by the
CMTS. The packet asks any new CM to announce itself on a specific upstream
channel.
CMTS sends a packet to the CM, defining its allocated downstream and upstream
channels.
CM then starts a process, called ranging, which determines the distance between
CM and CMTS. This process is required for synchronization between all CMs and
CMTSs for the minislots used for timesharing of the upstream channels.
CM sends a packet to the ISP, asking for the Internet address.
CM and CMTS then exchange some packets to establish security parameters,
which are needed for public network such as Cable TV.
CM sends its unique identifier to CMTS.
Upstream communication can start in the allocated upstream channel; CM can
contend for the minislots to send data.
Downstream communication: There is no contention because there is only one sender.
CMTS sends the packet with the address of receiving CM, using the allocated downstream channel.
McGraw-Hill
12
©The McGraw-Hill Companies, Inc., 2004
SONET Devices



McGraw-Hill
Fiber-optic cables are suitable for high data rate technologies like video
conferencing as well as to carry large numbers of lower-rate technologies at the
same time.
ANSI standard is Synchronous Optical Network (SONET) and ITU-T standard is
Synchronous Digital Hierarchy (SDH).
SONET:

A single clock is used to handle the timing of transmissions and equipment
across the entire network.

Network-wide synchronization adds a level of predictability to the system. This
predictability, coupled with a powerful frame design, enables individual
channels to be multiplexed, thereby improving speed and reducing cost.

SONET contains recommendations for the standardization of fiber-optic
transmission system (FOTS) equipment.

SONET physical specifications and frame design include mechanisms that
allow it to carry signals from incompatible tributary systems (DS-0 and DS-1).
This is a flexibility that gives SONET a reputation for universal connectivity.

SONET is a synchronous TDM system controlled by a master clock with a very
level of accuracy.

SONET bandwidth of the fiber is considered as one channel divided into time
slots to define subchannels.
13
©The McGraw-Hill Companies, Inc., 2004

SONET Devices: Three basic devices

STS Multiplexer/demultiplexer:


Regenerator:




STS regenerator is a repeater that takes a received optical signal and
regenerates it.
Regenerators in this system, however, add a function to those of physical layer
repeaters.
SONET regenerator replaces some of the existing overhead information
(header information) with new information. These devices function at the data
link layer.
Add/drop multiplexer:

McGraw-Hill
Either multiplexes signals from multiple sources into an STS or demultiplexes
an STS into different destination signals.
Can add signals coming from different sources into a given path or remove a
desired signal from a path and redirect it without demultiplexing the entire
signal.
14
©The McGraw-Hill Companies, Inc., 2004
Frame format





SONET frame can be viewed as matrix of nine rows of 90 octets,
for a total of 820 octets.
Some of the octets are used for control; they are not positioned at
the beginning or end of the frame (like a header or trailer).
First three columns of the frame are used for administration
overhead. The rest of the frame is called the Synchronous Payload
Envelope (SPE).
SPE contains transmission overhead and user data. The payload,
however, does not have to start at row 1, column 4; it can start
anywhere in the frame and can even span two frames. This feature
allows some flexibility; if the SPE arrives a little late, after a frame
has already started, the SPE does not have to wait for the
beginning of the next frame.
A pointer (address) occupying columns 1 to
3 of row 4 can determine the beginning
address (row and column) of the SPE.
McGraw-Hill
15
©The McGraw-Hill Companies, Inc., 2004
Frame Transmission




SONET frames are transmitted one after another without any gap in
between, even if there are no real data.
Empty frames carry dummy data. A sequence of frames looks like a
sequence of bits.
The first 2 bytes of each frame, called alignment bytes, F628 in
hexadecimal, define the beginning of each frame. The third byte is
the frame identification.
Synchronous Transport Signals




SONET defines a hierarchy of signaling levels called synchronous
transport signals (STSs).
Each STS level (STS-1 to STS-192) supports a certain rate, specified in
megabits per second.
The physical links defined to carry each level of STS are called optical
carriers (OCs). OC levels describe the conceptual and physical
specifications of the links required to support each level of signaling.
STS-1: STS-1 or OC-1 is the lowest-rate service provided by SONET.


McGraw-Hill
STS-1 transmits 8000 frames per second.
SPE bit rate is less than the raw bit rate due to three columns for
management.
16
©The McGraw-Hill Companies, Inc., 2004
Table 9.1 SONET rates
McGraw-Hill
STS
OC
Rate
(Mbps)
SPE (Mbps)
User
(Mbps)
STS-1
OC-1
51.84
50.12
49.536
STS-3
OC-3
155.52
150.336
148.608
STS-9
OC-9
466.56
451.008
445.824
STS-12
OC-12
622.08
601.344
594.432
STS-18
OC-18
933.12
902.016
891.648
STS-24
OC-24
1244.16
1202.688
1188.864
STS-36
OC-36
1866.23
1804.032
1783.296
STS-48
OC-48
2488.32
2405.376
2377.728
STS-192
OC-192
9953.28
9621.604
9510.912
17
©The McGraw-Hill Companies, Inc., 2004
Virtual Tributaries





SONET is designed to carry broadband payloads.
Current digital hierarchy data rates (DS-1 to DS-3) are lower than
STS-1.
To make SONET backward-compatible with the current hierarchy, its
frame design includes a system of virtual tributaries (VTs).
A virtual tributary is a partial payload that can be inserted into a frame
and combined with other partial payloads to fill out the frame.
Instead of using all 87 payload columns of a SPE frame for data from
one source, we can subdivide the SPE and call each component a VT.




VT1.5: The VT1.5 accommodates the U.S. DS-1 service (1.544 Mbps)
VT2: VT2 accommodates the European CEPT-1 service (2.048 Mbps)
VT3: VT3 accommodates the DS-1C
service (fractional DS-1, 3.152 Mbps)
VT6: VT6 accommodates the DS-2
service (6.312 Mbps)
McGraw-Hill
18
©The McGraw-Hill Companies, Inc., 2004

Virtual Tributaries [Contd..]


Higher-Rate Services


McGraw-Hill
When two or more tributaries are inserted into a single STS-1
frame, they are interleaved column by column. SONET provides
mechanisms for identifying each VT and separating them without
demultiplexing the entire stream.
Lower-rate STSs can be multiplied to make them compatible with higherrate systems.
For example, three STS-1’s can be combined into one STS-3, four STS3’s can be multiplexed into one STS-12, and so on.
19
©The McGraw-Hill Companies, Inc., 2004