Network Switching Types
Data can travel over several different types of lines using one of two
overall types of switching technology.
An example of a circuit switched network is a telephone network. The a remote access link for each major node in the circuit associated with a phone call is established for the duration of the call. It is also the case that the pathway taken might be different from one call to the next.
An example of a packet switching network is the standard routing operations on the Internet. When you send an e-mail it is broken down into packets. Each packet might take a different route to get to the destination. Then the packets are reassembled at the other end.
Circuit Switching Networks
This page surveys several circuit switching technologies in each of the following subsections. A lsit of these technologies follows.
POTS/PSTN. The POTS - Plain Old Telephone System or PSTN - Public Switched Telephone Network is the analog phone lines that are installed in most residences and businesses. It has two big advantages.
Using the existing telephone infrastructure to implement computer networking should make sense. Just about all that is required is a modem, to bridge the gap between digital computers and analog telephone lines, which are basically commodities.
Unfortunately, the original phone lines were not intended to carry the data rates required for many modern computer uses. This implies there are inherent limits to their capacity. Even with 56 Kbps modems, the capacity of many lines is 40 Kbps to 45 Kbps. The quality of the lines and transmissions is also a factor.
ISDN. ISDN - Integrated Services Digital Network was developed to replace the POTS - Plain Old Telephone Service and provide a reliable digital connection for both voice and data. The reality has proven to be much different. It hasn't replaced POTS and almost certainly won't because of the development of higher speed and still lower cost alternatives.
Some of the major characteristics of ISDN are contained in the following list.
ISDN is offered by the telephone companies in two standard access interfaces:
BRI is implemented for residential or small business data transfer. PRI is typically used for digital voice transmission in conjunction with PBX - Private Branch Exchange telephone systems.
DSL. DSL - Digital Subscriber Line is a newer technology than ISDN. It was developed as an add on service over existing copper wire lines. Some of the major characteristics of DSL are given in the following list:
DSL comes in several varieties:
xDSL is used to represent DSL in all of its possible implementations.
ADSL is based on the theory and experience that most customers require a much higher rate for downloads than uploads. Unfortunately, this will not be sufficient if you wish to have a web or FTP server. Most ADSL implementations make use of FDM - Frequency Division Multiplexing. Some make use of a different approach called echo cancellation. Echo cancellation is more efficient, but more expensive.
Leased Lines. If an organization wants to guarantee high performance and reliability they may end up choosing to lease lines from the telephone company. Using leased lines it is possible to get a direct connection from one location to another.
DDS. DDS is a digital service and was one of the first to be made publicly available. DDS provides a 56 kbps transfer rate. It has been superceded by T-carrier technologies because of their capabilities for higher capacity at lower cost.
T-Carriers. T-Carriers are dedicated digital circuits that can be leased to provide high speed data, voice and video over a relatively reliable point-to-point connection. These are usually implemented over copper wiring. They can also be implemented over fiber optic, coax and wireless technologies.
A CSU/DSU - Channel Service Unit/Digital Service Unit is used at each end of the connection to encode the data sent over a T-Carrier.
The T in T-Carrier refers to the transmission channel. The data signal transfer rate is referred to as the DS rate. Many people use an acronym such as DS-1 and T-1 to refer to the same type of connection. The following table outlines the most common T-Carrier implementations.
|Carrier Designation||Data Signal Rate||Data Transfer Speed||Number of
64 kbps Channels
|T-Carrier lines consist of 64 kbps channels. Thus
it is possible and often reasonable to lease part of a T-Carrier, called a
But, even though prices for these options have fallen dramatically, it is much cheaper to purchase DSL service than comparable T-1 service providing about the same bandwidth. Unfortunately, there aren't the same options for the higher bandwidth T-Carriers. On the other hand, you can purchase guaranteed bandwidth with the T-Carriers which still can make them more worthwhile.
Switched 56. Switched 56 is an enhanced version of the PSTN. It is a digital switched circuit connection which has a 56 kbps transfer rate over one channel. While inexpensive, its capacity is extremely limited. It is also a dialup technology.
Its biggest advantage over the PSTN is that it actually provides the 56 kbps that the PSTN doesn't really provide. Also, because it is digital it usually has lower error rates.
Both ISDN and DSL have proven to be more popular than Switched 56.
Packet Switching Networks
Packet switching networks are networks in which packets of data can take different routes to reach the same destination. At the receiving end the packets are put back together to reconstruct the original transmission. These sorts of networks are often depicted using a cloud because the exact routes of travel are unknown.
The three main types of packet switching technologies we will discuss are in the following list.
X.25. X.25 was one of the first packet switching networks. It was designed to work with IBM mainframes. It was originally called the ARPAnet 1822 protocol. The X.25 comes from these specifications established by the CCITT - International Telephone and Telegraph Consultative Committee. The CCITT changed its name to ITU - International Telecommunications Union in 1993.
X.25 is actually made up of several protocols operating at different layers of the OSI model.
Apparently, the primary objective in designing the X.25 protocol suite was reliability. At the time it was developed in the 1970s, both computers and telephone lines were more subject to errors. Thus, the PSDN running in X.25 includes redundant error checking. Thus reliability is improved, but transfer rates are slwoed by the extra error checking activity. The PSDN usually transfers at 64 kbps or less.
X.25 networks are still in use today. To make a WAN connection using this approach you can do one of the following.
Frame Relay. Frame Relay is an improvement of the older X.25 technology. It runs at only the two lowest layers of the OSI model, the physical and data link layers. It makes better use of newer more reliable telephone infrastructure and doesn't implement the extensive error checking of X.25. Frames with errors are discarded and the destination is required to detect missing packets and request retransmission.
Frame relay switches packets of a shared packet switching network owned by a regional telephone company such as MCI or AT&T. Frame Relays usually sit on top of T1 or T3 trunks operated by the provider. Long distance connections are over infrastructure owned by the frame relay provider. These connections are shared among a number of other customers.
Most customers rent PVCs - Permanent Virtual Circuits. These give a customer a continuous dedicated connection without paying for a leased line. Frame Relay customers are charged according to level of usage. It is usually used in the following ways
ATM. ATM - Asynchronous Transfer Mode was designed to support high speed applications such as video and/or audio streaming. ATM is actually specified as a Data Link Layer technology on layer 2 within the OSI Reference Model. ATM sends 53-byte cells instead of variable length packets. Using this fixed length results in certain advantages.
ATM is a switching technology in which virtual circuits are set up before a transmission starts. These can be either PVCs - Permanent Virtual Circuits or SVCs - Switched Virtual Circuits. A virtual circuit is built over a path of various hubs, switches and routers. Each end of the virtual circuit must agree to the path before the transmission can happen. This differs from Ethernet and Token Ring which transmit without prior notification or routing determination.
ATM is hardware based and the hardware, from NICs to hubs is costly. Standard transfer rates are
ATM can be used for both LANs and WANs. ATM is often associated with certain types of prioritizations to guarantee particular levels of QoS - Quality of Service. QoS involves controlling the allocation of network bandwidth for specific applications. ATM is heavily identified with multicasting of transmissions that have latency and/or priority sensitivities. In general, its fixed cell format and virtual circuits are much better at dealing with these kinds of issues.
Because of these approaches there are also a fair number of issues to resolve when using ATM as a backbone or WAN link because it must interact with other approaches.
ATM is designed to run over fiber optic cable operating the SONET - Synchronous Optical Network specification. We will discuss OC-SONET in the next section.
OC-SONET. SONET - Synchronous Optical Network specifications are set up for various cable speeds or OC - optical carrier levels. The following table displays the standard speeds.
51.84 Mbps fiber optic cable
|ATM is usually implemented on OC-3 or OC-12. Most
intercity links run OC-12, although some major backbone providers are now
Broadband ISDN. BISDN - Broadband ISDN is a technology designed to make use of fiber optic cable and radio waves to transmit data at high speeds over SONET, FDDI and Frame Relay. It is a broadband approach that can transmit multiple channels of voice, data and/or video over the same medium.
SMDS. SMDS - Switched Multiple Data Service is a packet switching technology that has been designed especially for WANs that experience bursts of traffic. It is connectionless, so it doesn't make use of circuits like some other approaches. It uses relatively large packets of up to 7168 bytes. Data transfer speeds are usually in the range of 1.544 Mbps up to 45 Mbps.
SMDS makes use of its own addressing scheme which uses 10 digit numbers to identify SMDS subnetworks. SMDS links are typically connected to SMDS switches on a telephone company's network. These are usually done on SONET OC-3 links.
SMDS was designed for public networks to provide LAN type services with the reach of a MAN. Unfortunately, it isn't that widely available and the equipment can be difficult to find.