Various Switch Architectures

On completion of this period, you would be able to know about

•           Network Hub

•           Network Switch

•           Store and Forward switch

•           Cut through switch

•           Adaptive switching mechanism

Network Hub

•           Hubs, switches, and routers are all devices used to connect one or more computers to other computers, networked devices, or to other networks

•            Each has two or more sockets called ports into which one can plug in the cables to make the connection 

•            A hub is typically the least expensive, least intelligent, and least complicated of the three

•            Hub takes the signal from  one port and sends out to the other        ports after due amplification and pulse shaping

•            Hub is a physical layer device

•             Hub is a multi-port repeater

HUB

Network Switch

•           A switch does essentially what a hub does but more efficiently

•            Switch also performs amplification, pulse reshaping functions like hub

•            Switch learns about the configuration of the   network

•            By observing the traffic, it learns which physical address computer is to which port

•            From the observation it prepares a forward table

•            When it receives a packet from a port, it reads the MAC address from the frame header and looks into forward table. Then forwards the packet only to the  port to which the machine with the destination MAC address is attached

•            Therefore Switch works both on physical layer and MAC layer

Network Switch

Network Switch

Various Switch Architectures

• Store and Forward 

• Cut through

• Adaptive 

Store and Forward  Switch

• A switch performing store-and-forward will wait to forward a frame until it has received the entire frame

• Store-and-forward is most often used in environments supporting reliable physical or data link protocols

• A received frame is often checked for errors before being forwarded

• This type of switch is inherently slower in environments 

•   where upper layer protocols already provide reliable          

•   services 

•  The tip off that you are dealing with a store and forward  

•    style switch is whether the switch has buffers

STORE AND FORWARD

Store and forward is a telecommunications technique in which information is sent to an intermediate station where it is kept and sent at a later time to the final destination or to another intermediate station. The intermediate station, or node in a networking context, verifies the integrity of the message before forwarding it. In general, this technique is used in networks with intermittent connectivity, especially in the wilderness or environments requiring high mobility. It may also be preferable in situations when there are long delays in transmission and variable and high error rates, or if a direct, end-to-end connection is not available.

This technique originates the delay-tolerant networks. No real-time services are available for these kinds of networks.

Store and forward networks predate the use of computers. Point-to-point teleprinter equipment was used to send messages which were stored at the receiving end on punched paper tape at a relay center. A human operator at the center removed the message tape from the receiving machine, read the addressing information, and then sent it toward its destination on appropriate outbound point-to-point teleprinter link. If the outbound link was in use, the operator placed the message in tape in a physical queue, usually consisting of a set of clips or hooks. A major relay center in the mid 1900s might have dozens of inbound and outbound teleprinters, scores of operators, and thousands of messages in the queues during peak periods. Operators referred to these centers as "torn-tape relay centers," a reference to removing the received message from the inbound teleprinter by tearing

the paper tape to separate one message from the next. The U.S. military term for such a center 

was "Non-Automated Relay Center" (NARC).

Cut Through Switch     

• Cut-through switches begin forwarding the frame as soon as the switch has read the destination address

• A cut through switch will forward the data before it has completed receiving the frame

• These switches will function at wire speed, forwarding traffic as fast as it receives it

• Nearly all cut-through switches have no RAM buffers for storing frames

In computer networking, cut-through switching is a method for packet switching systems, wherein the switch starts forwarding a frame (or packet) before the whole frame has been received, normally as soon as the destination address is processed. Compared to store and forward, this technique reduces latency(delaying the time) through the switch, but decreases reliability; corrupted frames are potentially forwarded.

Pure cut-through switching is only possible when the speed of the outgoing interface is equal to the incoming interface speed.

Use in Ethernet

When cut-through switching is used in Ethernet, because the frame check sequence appears at the end of the frame, the switch is not able to verify the integrity of an incoming packet before forwarding it. A cut-through switch will forward corrupted packets, whereas a store and forward switch will drop them.[1]

Fragment free is a variation on cut-through switching that partially addresses this problem by assuring that collision fragments are not forwarded. Fragment free will hold the frame until the first 64 bytes are read from the source to detect a collision before forwarding. This is only useful if there is a chance of a collision on the source port. 

Use in ATM

Cut-through switching was one of the important features of IP networks using ATM networks since the edge routers of the ATM network were able to use cell switching through the core of the network with low latency at all points. With higher speed links, this has become less of a problem since packet latency has become much smaller.

Use in InfiniBand

Cut-through switching is very popular in InfiniBand networks, since these are often deployed in environments where latency is a prime concern, such as supercomputer clusters.

-------------------------------------------------------------------------------------------------------------------------------

                                                     Adaptive Switch

• Adaptive switch is an advanced switching architecture

• It provides the communications infrastructure with unprecedented high availability, scalability

• It enables Peripheral Component Interconnect (PCI)-based telecommunications, data communications and embedded systems to incorporate 224 PCI bus  segments

Adaptive switching dynamically selects between cut-through and store and forward behaviors based on current network conditions

An adaptive switch is designed to operate in cut-through mode (cut-through switching) normally but if a port's error rate jumps too high, the switch automatically reconfigures the port to run in store-and-forward mode

This optimizes the switch's performance by providing higher speed cut-through switching switching if error rates are low but higher throughput store and forward switching if error rates are high.

Adaptive switching is typically done on a port-by-port basis.

In this class, you have learnt about

• Hub is a physical layer device

• Switch is a layer-2 device

• Store and Forward switch stores the entire frame in a buffer then it reads the header then forwards the frame to the concerned port

Cut through switch waits until it receives the frame header, then it forwards the frame to the concerned port without actually  waiting for the arrival of the complete frame