Introduction to WAN Technologies
Identify the basic characteristics (for example, speed, capacity, media) of the following WAN technologies:
Packet switching versus circuit switching
ISDN
FDDI
ATM
Frame Relay
SONET/SDH
T1/E1
T3/E3
OC-x
We'll start our examination of WANs by looking at perhaps the most significant consideration facing anyone who is implementing a WAN: whether to use a public or a private network as a means of connectivity. Both the private and public WAN network services have good and bad points, and knowing the difference between them is a good place to start discussing the characteristics of WANs.
Public Networks
The bottom line for many of the decisions made in networking often has to do with money. This is often true when choosing a WAN networking method. To save money and a certain amount of administrative effort, you can choose to set up a WAN using an existing transmission infrastructure. Two key public networks can be used to establish a WAN: the public switched telephone network (PSTN) and the Internet. Each of these is discussed in the following sections.
The Public Switched Telephone Network (PSTN)
The PSTN, which is often called plain old telephone system (POTS), is the entire collection of interconnected telephone wires throughout the world. Discussions of the PSTN include all the equipment that goes into connecting two points together, such as the cable, the networking equipment, and the telephone exchanges.
EXAM TIP
Use PSTN to Save Money If financial cost is a major concern, PSTN is the method of choice for creating a WAN.
The modern PSTN is largely digital, with analog connections existing primarily between homes and the local phone exchanges. Modems are used to convert the computer system's digital signals to analog, so they can be sent out over the analog connection.
Using the PSTN to establish WAN connections is a popular choice, although the significant drawback is the limited transfer speeds. Transfer on the PSTN is limited to 56Kbps with a modem and 128Kbps with an Integrated Services Digital Network (ISDN) connection, and it's difficult to share large files or videoconferencing at such speeds. However, companies that need to send only small amounts of data remotely can use the PSTN as an inexpensive alternative for remote access, particularly when other resources such as the Internet are not available.
The Internet
The Internet has become very popular for establishing WAN connections. Using the Internet to provide remote access creates a cost-effective and reliable solution for interconnecting LANs. One of the most common methods of using the Internet for connecting LANs is through the use of virtual private networks (VPNs). Essentially, a VPN uses a public network, such as the Internet, to connect private networks. Unlike private networks, VPNs can be used on an as-needed basis. A connection can be established to a remote location and then dropped when no transmissions are required. Many organizations use VPNs as dedicated links that permanently connect private LANs. Figure 7.1 shows a VPN connection over a public network.
Figure 7.1. A VPN connecting two private LANs.
NOTE
Cable and DSL The increasing availability of cable and Digital Subscriber Line (DSL) services has meant that companies are increasingly looking toward these technologies as a means to establish VPN connections. Cable and DSL are particularly suited for such a purpose because they offer high speeds for comparatively low cost. They are also available 24/7 for an inclusive cost, which is a bonus over other methods such as ISDN, which is billed on a usage basis. Cable and DSL, in concert with VPNs, are making it possible for companies to establish low-cost, secure WAN links. Previously, many companies would not have been able to afford a solution that offers this kind of speed, availability, and security.
Public Networks: Advantages and Disadvantages
The biggest advantages of a public network such as the PSTN and the Internet are accessibility and availability. Public network access is everywhere, and perhaps more importantly, it is inexpensive. In addition, the technologies required to use public networks, such as VPNs, are moderately easy to configure and can be implemented in a short amount of time.
You are most likely to see public networks being utilized by small organizations, where the money for a private network is simply not available or needed. For many small organizations, the capabilities that the Internet provides are sufficient.
As you might have already surmised, there are some drawbacks in using a public network to interconnect LANs. First and foremost is security. When you establish a link over a public network, there is a risk that your data may be compromised by another user on that network. Technologies such as VPNs put a lot of emphasis on security measures such as encryption, which is aimed at reducing the security risk. However, if you are sending sensitive data over a public network, there is always a risk. Discussions about the degree of risk are best left to hackers, crackers, and security experts, and that debate is sure to go on for a long time.
In addition to the security risks, there are numerous other considerations concerning public lines, such as disconnections, logon troubles for modems, Internet failures, and a host of other likely and unlikely circumstances. Keep in mind that with a public network you are getting something for nothing—or at least very little—and you will have to make concessions. If you can't live with the drawbacks, you can always switch to a private network.
Private Networks
If we all had unlimited IT budgets, most of us would be using private networking to connect LANs together. Private networks provide a solid way to maintain connectivity between LANs, at least for those who can afford them.
A private network does not suffer from the same considerations of a public network. There are many technologies used to create private networks, and they vary in cost and implementation difficulty. The specific technologies used to create WANs over private networks are discussed in detail in this chapter; the most common private network technologies include Asynchronous Transfer Mode (ATM), Frame Relay, and X.25.
A private network can be designed and implemented from scratch based on an organization's specific needs. The network can be as complex or simple, expensive or inexpensive, secure or insecure as allowed by the budget, location, utilization, and data usage demands. The network can also be designed around the security needs of the data being carried over the network. For instance, fiber-based networks are more secure and more expensive than copper-based networks or wireless networks. A private network can employ various protocols based on security or performance needs. Basically, a private network gives the designer an opportunity to correct most of the problems presented by public networks.
Probably the biggest single disadvantage of a private network is the cost. Whereas the PSTN is yours for the asking at a nominal monthly fee, a private network requires that you purchase or lease every piece of cable, all the network cards, hubs, routers, switches, and so on, until you have enough equipment to go live.
Because of the required networking equipment, private networks often require more administrative effort than public networks, where the networking infrastructure is maintained by outside administrators. Often, a hidden cost associated with private networks is the need for qualified people to manage and maintain them. As the network grows—and it will—it becomes increasingly complex and requires more and more attention. Good administration is a must, or inefficiencies and lack of dependability will quickly consume the value of the private network. A company needs to carefully weigh administrative issues before getting into private networking.
Using a private WAN need not be a total do-it-yourself approach. In fact, most telephone companies provide managed WAN services, which include all the equipment you need to create a WAN. They also monitor and manage the connection for you, making sure that everything operates as it is supposed to. There is, of course, a price attached to such a service, but for many companies, a managed solution is money well spent.
Switching Methods
Before we go on to discuss the specific WAN technologies, we must first look at an important element of the WAN technologies—the switching methods. In order for systems to communicate on a network, there has to be a communication path between them on which the data can travel. To communicate with another entity, you need to establish a path that can be used to move the information from one location to another and back. This is the function of switching: It provides a path between two communication endpoints and switches the data, to make sure it follows the correct path. Three types of switching are used most often in networks today:
EXAM TIP
Know the Differences For the Network+ exam, you will be expected to identify the differences between the various switching methods.
Packet switching
Circuit switching
Packet Switching
In packet switching, messages are broken down into smaller pieces called packets. Each packet is assigned source, destination, and intermediate node addresses. Packets are required to have this information because they do not always use the same path or route to get to their intended destination. Referred to as independent routing, this is one of the advantages of packet switching. Independent routing allows for a better use of available bandwidth by letting packets travel different routes, to avoid high-traffic areas. Independent routing also allows packets to take an alternate route if a particular route is unavailable for some reason. Figure 7.2 shows how packets can travel in a packet-switching environment.
Figure 7.2. An example of packet switching.
In a packet-switching system, when packets are sent onto the network, the sending device is responsible for choosing the best path for the packet. This path might change in transit, and it is possible for the receiving device to receive the packets in a random or nonsequential order. When this happens, the receiving device waits until all the data packets are received, and then it reconstructs them according to their built-in sequence numbers.
NOTE
Packet Switching Packet switching is the most popular switching method for networks and is used on most LANs.
Two types of packet-switching methods are used on networks: virtual-circuit packet switching and datagram packet switching. Each of these methods is described in the following sections.
NOTE
Packet Size Restrictions The packet size is restricted in a packet-switching network, to ensure that the packets can be stored in RAM instead of on a hard disk. The benefit of this size restriction is faster access because retrieving data from RAM is faster than retrieving data from a hard disk.
Virtual-Circuit Packet Switching
When virtual-circuit switching is used, a logical connection is established between the source and the destination device. This logical connection is established when the sending device initiates a conversation with the receiving device. The logical communication path between the two devices can remain active for as long as the two devices are available or can be used to send packets once. After the sending process has completed, the line can be closed.
All the packets in virtual-circuit packet switching must follow the same path. That is, they travel through the logical communication path that is established. Virtual-circuit packet switching is commonly used for connection-oriented services such as real-time video.
Datagram Packet Switching
Unlike virtual-circuit packet switching, datagram packet switching does not establish a logical connection between the sending and transmitting devices. The packets in datagram packet switching are independently sent, meaning that they can take different paths through the network to reach their intended destination. To do this, each packet must be individually addressed, to determine where its source and destination are. This method ensures that packets take the easiest possible routes to their destination and avoid high-traffic areas.
Because in datagram packet switching the packets can take multiple paths to reach their destination, they can be received in a nonsequential order. The information contained within each packet header is used to reconstruct all the packets, and so the original message is received intact.
NOTE
Datagram Packet Sizes The data packet size used with datagram packet switching is kept small in case of error, which would cause the packets to be resent.
Circuit Switching
In contrast to the packet-switching method, circuit switching requires a dedicated physical connection between the sending and receiving devices. The most commonly used analogy to represent circuit switching is a telephone conversation, in which the parties involved have a dedicated link between them for the duration of the conversation. When either party disconnects, the circuit is broken and the data path is lost. This is an accurate representation of how circuit switching works with network and data transmissions. The sending system establishes a physical connection, the data is transmitted between the two, and when the transmission is complete, the channel is closed.
Some clear advantages to the circuit-switching technology make it well suited for certain applications. The primary advantage is that after a connection is established, there is a consistent and reliable connection between the sending and receiving device. This allows for transmissions at a guaranteed rate of transfer.
Like all technologies, circuit switching has downsides. As you might imagine, a dedicated communication line can be very inefficient. After the physical connection is established, it is unavailable to any other sessions until the transmission is complete. Again using the phone call analogy, this would be like a caller trying to reach another caller and getting a busy signal. Circuit switching can therefore be fraught with long connection delays. Figure 7.3 shows an example of circuit switching.
Message Switching
In some respects, message switching is similar to packet switching, but instead of using and sending packets, message switching divides data transmissions into messages. Like packets, each of these messages contains the destination address. Devices on the network that forward the message use this destination address. Each intermediate device in the message's path stores the message momentarily and then forwards it to the next device in the network, until it finally reaches its destination. Message switching is therefore often referred to as the store-and-forward method.
NOTE
Email and the Store-and-Forward Method The store-and-forward method used by message switching makes it well suited for certain applications, including email.
Message switching offers many advantages over circuit switching. Because message switching doesn't require a dedicated connection as does circuit switching, a larger number of devices can share the bandwidth of the network. A message-switching system also has the capability of storing messages, which allows the traffic on the network to clear. This strategy can significantly reduce the traffic congestion on the network.
The main drawbacks of the message-switching method are that the store-and-forward method makes it a poor choice for real-time applications, such as videoconferencing, in which the temporary storing of data would be disruptive to the message. A second drawback is that the intermediate devices, often PC systems, must be capable of temporarily storing messages by using their hard disk space.
REVIEW BREAK: Switching Methods Comparison
Table 7.1 summarizes the characteristics of the various switching methods.