ATM Network Topology
Asynchronous transfer mode (ATM) is the underlying technology enabling broadband integrated services digital network (B-ISDN). B-ISDN was introduced as the successor to narrowband ISDN after the latter fell short of meeting the high demand for bandwidth required by emerging applications such as real-time video and high definition TV (HDTV). B-ISDN envisions the transmission of fixed-size packets (cells) over digital virtual circuits at rates exceeding 150 Mb/s.
ATM is basically a packet-switching technology with 53-byte-long cells. The small cell size makes it possible to build switches that can accept and switch a large batch of cells. ATM is asynchronous in that it has no requirements that cells rigidly alternate among the various sources (i.e., cells arrive randomly from different sources). The ATM standards set by the forum define the user-network interface; that is the way a computer owned by a private user can connect to the network and communicate through it.
There are five different categories of services are used for different types of communication:
· Constant Bit Rate (CBR)
It is designed to support real-time applications, like video and audio. Peek cell rate (CBR) is used by connections that require a fixed amount of bandwidth, characterized by a PCR value that is continuously available during the connection lifetime.
· Variable Bit Rate (VBR)
This category is used allows users to send at a variable rate. It is designed to support burst traffic (either video or frame relay traffic). This category is subdivided into two categories: Real Time VBR like interactive compressed video and Non real Time VBR like multimedia email .
· Available Bit Rate (ABR)
In this service category, the cell rate depends on the availability of the network. It is basically designed for bursty traffic whose bandwidth range is known roughly . It is intended for sources that have the ability to reduce or increase their information rate if the network requires them to do so. ABR guarantees only a minimum amount of bandwidth and may be limited to a specified peak emission rate .
· Unspecified Bit Rate (UBR)
This category uses the left-over network capacity . It is a "best effort" service intended for non-critical applications which do not require tightly constrained delay and delay variation or a specified quality of service. UBR service shares the remaining bandwidth without any specific feedback mechanisms. .
· Guaranteed Frame Rate (GFR)
This service category is intended to support non-real-time applications requiring a minimum rate guarantee. It does not require adherence to a rate control protocol. An example application is frame relay inter working .
ATM network with bottleneck
The ATM Forum adopted a rate-based congestion control scheme for the ABR service . In this scheme, explicit rate control messages are sent from intermediate nodes to the sources using special cells called resource management (RM) cells. The goal of this congestion control mechanism is to fairly share the band with left over from high-priority traffic (CBR and VBR) among the ABR sources while making sure that the links throughout the network are fully utilized.
An ATM network consists of several nodes (switches) interconnected via bidirectional links. We say that a switch is bottlenecked if the incoming ABR cell rate at any one of its output ports is larger than the available rate to serve the ABR cells. Clearly, at a given instant, there may be multiple bottlenecks in the network. As shown in figure below.
Bottleneck in the network
Although, the rate-based congestion control schemes are standardized, developing good explicit rate computation algorithms is still an open issue. As the link speeds continue to rise the delay-bandwidth product (i.e., the product of the round-trip propagation delay and the link capacity) increases.
Flow Control Model for ABR
In the ABR service, the source adapts its rate to changing network conditions. ABR flow control occurs between a source and destination, which are connected via bidirectional links. The forward direction is the direction from the source to the destination, and the backward direction is the direction from the destination to the source.
A source generates forward RM cells every N r m data cells, where N r m is generally taken to be 32. These cells travel along the same path as the data cells but are treated specially by the switches along the way. The switch may:
A Schematic of how ABR works
On establishment of an ABR connection, the source specifies to the network both a maximum required bandwidth and a minimum usable bandwidth. These are designated as peak cell rate (PCR) and the minimum cell rate (MCR), respectively. The MCR may be specified as zero. The bandwidth available from the network may vary, but should no become less than MCR. Each ABR source has a current cell rate, allowed cell rate (ACR), which it must modify upon receiving feedback from the network via RM cells.
The ACR always falls somewhere between MCR and PCR. When a source sends out a forward RM cell, it sets the ER field of the RM cell to the rate at which it would currently like to transmit. As the RM cell passes through the various switches on the way to the destination and back to the source, those that are congested may reduce the ER. The following figure is showen some fields and their positions in RM-cells:
The Network Model
The Available Bit Rate service (ABR) is a "best effort" service intended for traffic which imposes no bound on delay or delay variation. The network guarantees a Minimum Cell Rate (MCR) for an ABR source, which is negotiated at the connection set up, and commits to fairly divide the unused bandwidth among all ABR sources.
Consider a graph of communication links shared by a number of connections (source-destination pairs).
Let denote the set of links in the network.
Let denote the set of connections using these links.
, denote the each connection with flow rate.
, used to designate the set of links used by s connections.
, used to design the set of connections using link.
Each link has a capacity denoted and a finite size buffer.
, denoted by minimum rate.
By assuming that each source, either controlled or uncontrolled, has an MCR=0. If the MCRs are positive, we can reserve the minimum cell rate for each user and make the assumption that a source will attempt to send at a rate no smaller than its reserved MCR.
, this corresponds to the rate in excess of.