RSS
Twitter
Facebook
Showing posts with label ATM. Show all posts
ATM × COMPUTER NETWORKS-I
The purpose of the ATM Adaptation
Layer (AAL) is to adapt the PDUs passed down from the higher layer onto ATM
cells. As the higher level PDUs may in general be of an arbitrary size, so one
of the two sublayers in the AAL is responsible for segmentation and
reassembly (SAR) of the higher layer PDUs. The other sublayer, the convergence
sublayer (CS), is responsible for packaging the higher layer PDU with any
additional information required for the adaptation necessary and offering an
interface to the B-ISDN user. As mentioned previously, there are many and
varied requirements for applications using the B-ISDN and as there are several
AAL protocols defined. The classification of these protocols was guided by
considering the following parameters:
· Timing
relationship. The
(non-)requirement for synchronisation between the
receiver and sender.
· Bit rate. Constant or
variable.
· Connection mode.
Connection
oriented or connectionless.
A summary of the (sensible)
combinations of these parameters is given in Figure . Each of these classes may
be looseley associated with the the ATM Forum classes described previously:
Class A, CBR; Class B, VBR; Class C, ABR(ish); Class D, UBR. This mapping also
give example use of the various classes. The various AAL classes are provided
by several AAL protocols identified by a type number. The use of a particular AAL
type for the provision of a particular service is not wholy well defined.
Figure: AAL service
classification
AAL Type 0. This is
effectively a NULL AAL. It is not really an official AAL type but is mentioned
for completeness.
AAL Type 1. This AAL type is
normally used by Class A (CBR) services. The function performed by this AAL
are:
· Segmentation and
reassembly of user information.
· Handling of cell
delay variation (jitter).
· Handling of cell
reassembly variation.
· Handling of lost
and misinserted cells.
· Source clock
frequency recovery at receiver.
· Source data
structure recovery at receiver.
· Monitoring and
handling of AAL-PCI bit errors.
· Monitoring and
(possibly) correcting the bit errors in the user information field.
· For circuit
emulation, monitoring and maintenance of end-to-end QoS.
AAL Type 2. This AAL type
would be used with Class B (VBR). This type is not well defined and it seems
possible that the it may be merged with AAL Type 1 in the future. Some of its
functions are similar to ALL Type 1:
· Segmentation and
reassembly of user information.
· Handling of cell
delay variation (jitter).
· Handling of lost
and misinserted cells.
· Source clock
frequency recovery at receiver.
· Monitoring and
handling of AAL-PCI bit errors.
· Monitoring and
(possibly) correcting the bit errors in the user information field.
It also has the additional
functions:
· Handle SDUs from
a variable bit rate source.
· Transfer timing
information between source and destination.
· Notify the
higher layers of uncorrectable errors in AAL.
AAL Type 3/4. There was once
separate Type 3 and Type 4 AALs, but they have now been merged. This AAL is now
intended to support both Class C (ABR) and Class D (UBR) services.
In this AAL, the convergence
sublayer is split into two (Figure ), the service specific convergence
subslayer (SSCS) and the common part convergence sublayer (CPCS). The
SSCS is application dependent, i.e. it could be for a VBR video application.
The CPCS is responsible for constructing PDUs that can be sent to the other end
user. There are two modes of operation of ALL Type 3/4; message mode and
streaming mode.
Figure: AAL Type 3/4
sub-layering
The message mode is intended for
use framed data where the AAL-SDU is a logical unit of data with respect to the
B-ISDN user (Figure ). It allows the the transport of a single AAL-SDU in one
or (optionally) more than one CS-PDU. The CS-PDU may be then further spilt into
several SAR-PDUs. The AAL-SDU can be of an arbitrary size.
Figure: AAL Type 3/4
message mode service
In streaming mode, the AAL-SDUs
are of fixed size and one or more of them may be transported in a single CS-PDU
(Figure ). Each AAL-SDU is delivered in a separate SAR-PDU.
Figure: AAL Type 3/4
streaming mode service
In both cases, the SAR sublayer
provides error detection and both these modes can offer the following operational
procedures:
· Assured
operation. Flow
control and retransmission of missing or erroneous
AAL-SDUs. Flow control restricted
to point-to-point connections at the ATM
layer and point-to-multipoint
flow control possible.
· Non-assured
operation. No
retransmission of missing or erroneous SAR-PDUs.
Optionally deliver erroneous PDUs
to user. Allow flow control for point-to-point
connections but not
point-to-multipoint.
This AAL type also provides
multiplexing at the SAR sublayer. AAL Type 5.
This AAL type provides similar
services as AAL Type 3/4, but has a reduced overhead when compared to AAL Type
3/4. It is intended for use by VBR sources with timing relationship between
source and destination. It has identical modes and operational procedures as
AAL Type 3/4. The difference is that this AAL does not provide the AAL Type 3/4
multiplexing capability. The SAR sublayer accepts only AAL-SDUs that are an integer
multiple of 48 octets. So, it would be possible for this AAL to offer an
efficient cell based interface to the B-ISDN user.
AAL Types 1 and Type 2 are not
used much.
AAL Type 3/4 has been chosen for
the provision of the Bellcore switched multimegabit data service (SMDS),
the European version of which is the connectionless broadband data service
(CBDS).
AAL Type 5 has been selected by
the IETF to provide IP services over ATM.
ATM × COMPUTER NETWORKS-I
To provide the new B-ISDN
services, use of a technology called asynchronous transfer mode (ATM) is
specified by ITU. ATM is a connection-oriented technique based on the use of
fixed size packets termed cells. These cells are 53 octets in size, with
5 octets used for the cell header, and the remaining 48 octets for data
Figure: An ATM cell
The term asynchronous transfer
mode needs some explanation. The words transfer mode say that this
technology is specific way of transmitting and switching through the network.
The term asynchronous refers to the fact that the packets are
transmitted using asynchronous techniques, and the two end-points need not have
synchronised clocks. Also, the use and allocation of cells and their subsequent
multiplexing and transmission through the network is determined in an
asynchronous fashion, e.g. on demand, and is independent of the user. ATM will
support both circuit switched and packet switched (sometimes referred to as circuit
mode and packet mode, respectively) services.
Figure: The principle of
asynchronous transfer mode
There is much to know about ATM,
however, here we will consider only the transmission aspects, by looking at the
Physical layer and the ATM layer. There are several types of cells to consider,
some are passed all the way through to the
ATM layer, while others never
proceed beyond the Physical layer:
· Assigned Cell. (ATM Layer) A
cell which is used to provide the service to the
higher
layers. This is the cell type that is used in transporting the higher level
PDU.
· Unassigned cell. (ATM Layer) An
ATM layer cell which is not an assigned cell.
· Valid cell. (Physical layer)
Any cell that does not contain header errors, either
through
successful transmission or after being corrected by the physical layer.
· Invalid cell. (Physical layer)
A cell with an error in its header that can not be
corrected.
It will be discarded by the physical layer.
· Idle cell. (Physical layer)
A cell that is inserted/extracted by the physical layer for the purposes of
payload capacity adaptation.
As explained earlier, there are
two interfaces to the B-ISDN to consider -- the user-tonetwork interface (UNI)
and the network-to-network interface (NNI). For these two interfaces the
internal structure of the ATM cell header is slightly different, as shown in Figure
The various parts of the cell
header are:
· Generic flow
control (GFC). (4
bits) Default value 0000. The exact use of this
field is, as yet, not fully
defined. This is used only in assigned or unassigned (i.e.
ATM layer cells).
· Virtual path
identifier (VPI). (8
or 12 bits) This is used for physical layer
routing, together with the;
· Virtual circuit
identifier (VCI). (16
bits) This is used in conjunction with the
VPI field to provide physical
level routing. Many cells with different VCIs may
have the same VPI. Some VCI/VPI
values are pre-assigned for special purposes
· Payload type
(PT). (3
bits) Some PT values -- PT identifiers (PTIs) -- are preassigned for
special purposes.
· Cell loss
priority (CLP). (1
bit) Flag stating whether or not this cell can be
dropped in the prescence of
network congestion or other network difficulties. A
value of 0 means that this cell
has high priority and should not be discarded.
· Header error
control (HEC). (8
bits) This is used by the physical layer for
detecting and correcting errors
in the cell header.
ATM × COMPUTER NETWORKS-I
Why
ATM?
•All digital
telephone network
•Need for data
applications like computer communications and facsimile
•Nonvoice
applications like videoconferencing need to be included in the future networks
•Circuit switching
not suitable but packet switching would accommodate this
•ISDN were the first
effort to address this
ISDN Overview
•There are many types
of WAN technologies that can be used to solve the problems of users who need
network access from remote locations.
•ISDN has been
specifically designed to solve the low bandwidth problems
that small offices or dial-in users have with traditional telephone dial-in
services.
•Telephone companies
developed ISDN with the intention of creating a totally digital
network whilst making use of the existing telephone wiring system.
•ISDN works very much
like a telephone - When you make a data call with ISDN, the WAN link is brought
up for the duration of the call and is taken down when the call is completed
•ISDN allows digital
signals to be transmitted over existing telephone wiring.
••This became
possible when the telephone company switches were
upgraded to handle digital signals.
••ISDN is generally
viewed as an alternative to leased lines,
which can be used for telecommuting and networking small and remote offices into
LANs.
•ISDN's ability to
bring digital connectivity to local sites has many benefits, including the following:
••ISDN can carry a variety
of user traffic signals including:
–digital video,
packet-switched data, and telephone network services.
•ISDN offers much faster
call setup than modem connections because it uses out-ofband (D, or
delta, channel) signaling.
–For example, some
ISDN calls can be setup in less than one second.
•ISDN provides a faster
data transfer rate than modems by using the bearer channel (B channel of
64kbps).
–With multiple B
channels, ISDN offers users more bandwidth on WANs than some leased lines.
–For example, if you
were to use two B channels, the bandwidth capability is 128Kbps because each B
channel handles 64Kbps.
–•ISDN can provide a clear
data path over which to negotiate PPP links.
BRI
(Basic Rate Interface)
Connection from the
ISDN office to the user location provides for access to three
channels. The channels are two 64Kb B-channels and one 16Kb D-channel
The B-channels and
the D-channel provide the user with access to the circuit switched network
PRI
(Primary Rate Interface)
ISDN Primary Rate
Interface service provides digital access via a T1 line. A T1 line provides a
1.544 bandwidth. This bandwidth is divided into 24 64Kb channels. The ISDN PRI
service uses 23 B channel access and uses the 24th
(D) channel for signaling purposes
BSDN
Reference Model
•There are three
planes
•User Plane
•Control Plane
•Management Plane
User Plane
•ATM is specified via
a three-layer reference model:
•Physical layer (OSIs
physical layer)
•ATM layer (generally
OSIs data link layer)
•ATM adaptation layer
(AAL) (generally OSIs higher-level layers (transport, session, and application)
