We will explain the functions and correct uses of the WAN and describe the protocols that to get involved and how they map to the OSI model. In listing the components, we will describe hardware devices with routers, playing a major role in WAN infra structures. Finally, we will try to describe and list the layer 2 protocols commonly used in WAN deployments.
A wide area network is a data communications network that operates beyond the geographic scope of a LAN. There are three major characteristics of LANs. They connect devices that are separated by wide geographical areas. They also use the service of carriers such as cell phone companies or cable companies, satellite systems, and network providers. Typically, the customer equipment will use various types of serial connections to connect to the wide area network.
The Internet could be seen as the WAN of whims, it covers all requirements and major characteristics, however, the term WAN is often referred to and used in private networking scenarios. In other words, for connectivity between offices and branches of the same organization. To the service provider, it is the backbone of sources of revenue not only for connectivity to customers but also for additional services like internet access, office-to-office connectivity, and voice transmissions among others.
Need for WANs
So WANs were born to meet the wide area communications needs of organizations of all kinds. Through the years though, the concept has been expanded to not only connect branch offices to remote offices, but also allow organizations to communicate with business partners, suppliers and customers. Also, with the transient mobility and universal access, telecommuters and mobile workers have been included as beneficiaries of WAN services, due to the pneumatic nature of those connections. More ubiquitous networks like the Internet have been used to expand the WAN and allow connectivity to this mobile users.
WANs vs. LANs
The differences between WANs and LANs are intuitive. One is for wide area connectivity; the other one is for local connectivity within a small geographic area, buildings, campuses, etc. Also, the WAN typically uses an outside service provider, whereas the LAN is owned by the organization. The important point in today’s reality though is the fact that the boundaries are blurring and if we think of technologies like MPLS and virtual private networks, then a wide area connection is virtually part of the LAN and a logical extension of the LAN and is treated other than distance and perhaps performance as another LAN connection.
There is also the concept of the metropolitan area network (MAN), which is perhaps a smaller WAN with some LAN capabilities, if you think of metro Ethernet services at a high speed. In terms of this criteria though, area and ownership, there is a clear distinction between WANs and LANs.
WAN Access and the OSI Reference Model
When your organization connects to a wide area service provider, the conversation will be typically along the lines of physical layer and data link layer. The service provider will define the physical layer options and of course that deals with the electrical, mechanical, and operational features of the connection. Access options to that media will also be defined and some options are listed here, frame-relay is one, ATM or HDLC encapsulation on serial links.
In terms of components at both layers and more, the WAN connection typically looks like this:
A router at a customer premises providing traffic segmentation and also the wealth of interfaces that can connect to service provider offerings. Today’s routers include modular chassis that can upgrade or change to a different service used by changing the modules or network cards. The router will be typically connected to a modem or DSU/CSU, depending on the type of service, and those devices will be responsible for converting the signals coming from the router into whichever format the service provider transmission requires. We can define the following devices:
- WAN switches
- Core routers
Also, those devices sometimes define the demarcation point between the administrative scope of the provider and the administrative scope of the customer. The cloud there represents the service provider network and those technologies listed here: ATM, frame relay, or the PSTN are implemented as a full network of WAN networking devices with high levels of reliability and availability to support multiple customers. Perhaps the one option not listed here is MPLS or Multi-Protocol Label Switching.
If we use a magnifying glass to look at the customer premises and the required equipment there, we will see a data terminal equipment, typically a router connecting to a data communications equipment, which prepares the data sent by the router for transmission into the service provider network.
WANs have been around for a while and the people who build these WAN technologies like to come up with a lot of acronyms. And so, we're going to see some extra acronyms that we may not be super used to yet. CPE, what does CPE stand for? It stands for Customer Premise Equipment.
This means, there is gear on our site. So that's not a router that is sitting on the street that connects to us, no, that's inside of our building. There is also a term that we don't see here called demarc – also known as Demarcation Point – where stuff on our side of demarc is our responsibility, and on their side of the demarc, it's the service provider's responsibility. So, we see some other acronyms here, oh, CSU/DSU, that's a big one, and something we definitely want to understand what it's about. So this channel service unit/data service unit, what does that do for us?
Well this device is going to be responsible for terminating that leased line, specifically, a digital leased line from our ISP or service provider. An example would be a T1 connection.
Let's talk about a T1 connection. We're not going to really focus too much on it, but a T1 connection is really interesting. First, how fast is a T1? 1.544 megabits per second. Okay, and also if you're in Europe, E1 would be 2.048 megabits per second. If we were to dissect this, we get out our scalpel and you know, the formaldehyde, and we dissect a T1 or an E1 or an N1, what we would find is a channelized service. Channelized being, we've got more than one sort of pathway inside of that for the signalling. And that's why we need this device. It's going to terminate a digital line like a T1 or an E1. And we need it because we've got this thing that our router isn't prepared for. The router is not prepared for these 24 channels. So that's the thing that makes these 24 channels into one usable data circuit that we can latch onto. That's a pretty important thing.
How are we going to connect that router up to the CSU/DSU? We have a serial cable. But notice that there is another term here associated with this CSU/DSU. What is it called? DCE – also known as data communications equipment.
Yes, DCE, that's a general term, just so we're clear at this point. DTE – also known as data terminal equipment – and DCE are general terms that apply to not just this. CSU/DSU, that's a very specific thing. DCE is one side of that serial cable, so I want you to focus here. We were pointing at the serial cable that we see here, there are two sides of the serial cable, data terminal equipment or sometimes said data terminating equipment, that's fine. And then, data communications equipment, it is usually what most people say the CE is, data communications equipment. Two sides of a serial connection which are different. This is fundamentally different than the cable types that we've learned about with Ethernet.
Ethernet doesn't really matter which side is which. But does it matter to get the cable right when we're using a serial cable, does it matter DTE goes on one side, DCE goes on another?
It definitely matters and we have to make sure we cable it correctly. Because if we don't cable it correctly, we're not going to experience the communication that we expect. So we'll see a little more on this later and it's really, really important. And honestly, knowing the serial cable is the most important thing that you can pick up in this discussion in this lesson.
If you're talking broadband, that could be cable, it could be DSL – also known as digital subscriber line. And think about what you use in your home. You probably you have DSL or cable in your home, okay, right? Think about that.
You've got your broadband router, I don't care what vendor it is. It's probably not Cisco and you're connecting that up to some external box that could be a DSL modem, it could be a cable modem. And what does modem stand for? Modulate and demodulate. And they are modulating and demodulating, even though this isn't a modem for a regular telephone line, which would be an asynchronous modem, it would be an analog modem. These are some modems that connect up to the telephone line, that's DSL. And if this was a cable modem, it would connect up to coaxial pipe that we get our TV service through, and that cable modem does quite a bit, or the DSL. But it gives us connectivity, and we use Ethernet typically to cable ourselves to those modems.
We see also, there's potential for an optical converter here. Fiber optic to your sites, is a little less common in the customer side of things. But it certainly has been around a while on the production enterprise network side of things. Fiber is great, because it can go really far, that's its real strength. The photons that travel in the fiber-optic core don't have to deal with resistance like with all the others that are based on sending electric signals. So this is very different signaling type and we need something to terminate that. And in all of these, I would classify the cable modem and the optical converter also as DCEs from a general classification standpoint, right? They are the things that we attach to, to get connectivity. They are that intermediary. So even a classic regular modem would also be classified as a DCE when we're using the general terminology. But understand also, there's a specific terminology that applies that to serial cables, which is very important.
We mentioned a very important term earlier, the demarc point. Folks, look at the devices we have here. We have routers, CSU/DSU, DSL modems, optical converters, we also have the Internet. Where would you typically find that demarc point? Now that we've defined all these devices for you, where would you typically find that demarc point? What is that point all about? It's that dividing line between where your responsibility begins and your service provider's ends, so it's that connection. So if you're saying CSU/DSU or DSL modem or optical converter, that would be that dividing line, that would be the demarc point. So the cables coming in, whether it's the telephone, the coaxial cable, the fiber cable, whatever it is, digital, analog from the ISP connects up to these devices or from that point on inwards, it's your responsibility. So would you own these DCEs as we call them or would you lease them? It could be either, or. You could own them or you could lease them, it really all depends on your service provider and the type of contract you have with them.
Serial Point-to-Point Connections
The serial cable connecting the DTE and the DCE will have different options to terminate at both devices, a router and the CSU/DSU, in this example.
Typically, Cisco routers support a 60-pin connector on the router side. More recent models support the 26-pin connector, which is less bulky and supports higher densities. On the DCE side, several options are available including V35, X21, and RS 232, with differences in terms of distance and speed.
WAN - Multiple LANs
When you look at the bigger picture, what we have is a connection of LANs connected by the wide area network. Routers will connect to the LAN and optionally you offer services like DHCP, inter-VLAN routing, and more recently things like voice services acting as a gateway and offering call routing capabilities. The WAN connection will be dependent on the flavor of wide area network service and the different offerings by the service provider. The router will provide routing functions to forward packets to those remote destinations and nowadays it also provides security options to filter traffic, keep the "bad traffic out" and also actually provide firewalling services.
WAN Data-Link Protocols
The WAN will also provide data link layer capabilities for access to the media. The protocols and encapsulation methods are here:
- Frame Relay (LAPF)
All of them are layer 2 and they go from the more simple scenario of a point-to-point link, encapsulating packets with HDLC or PPP to the more complex scenarios based on virtual circuits like frame relay and ATM. All of these include encapsulation and framing capabilities and will present different options and advantages and disadvantages that define their use.
WAN Link Options
In choosing, a wide area network service, it is important to understand the categories and their advantages and disadvantages. One of the criteria to compare is cost versus availability and bandwidth, and so dedicated lines will be exactly that. Dedicated point-to-point links typically are leased from a carrier. That is why they are called leased lines, and their use is typically linked to the willingness of users to pay for these dedicated lines. Switched options are typically shared networks that will be less costly in terms of data communications.
Circuit switch connections are typically used for voice and the main example is the public switched telephone network, and establish a virtual connection for the voice or data between sender and receiver. Other options also create virtual connections across the shared network. In the case of frame relay, you have virtual circuits to transmit packets.
In the case of ATM, you have smaller packets called cells and the network is better equipped to prioritize traffic and transmit multimedia, voice, data, and some others. Cable and DSL are more recent packet switched networks, typically used for Internet access.
In summary, the wide area network expands the LAN and breaks geographic barriers. Today’s WANs have to consider mobile users and traveling workers.