LANs and WANs (7.1.1)

Whether at work or at home, we all use local-area networks (LANs). However, LANs are limited to a small geographic area. A wide-area network (WAN) is required to connect beyond the boundaries of a LAN. A WAN is a telecommunications network that spans a relatively large geographic area and operates beyond the geographic scope of a LAN.

In Figure 7-1, WAN services are required to interconnect an enterprise campus network to remote LANs at branch sites, telecommuter sites, and remote users.

Table 7-1 highlights the differences between LANs and WANs.

Private and Public WANs (7.1.2)

WANs may be built by a variety of different types of organizations, including the following:

• An organization that wants to connect users in different locations

• An ISP that wants to connect customers to the internet

• An ISP or telecommunications provider that wants to interconnect ISPs

A private WAN connection is dedicated to a single customer. It provides for the following:

• Guaranteed service level

• Consistent bandwidth

• Security

A public WAN connection is typically provided by an ISP or a telecommunications service provider using the internet. With a public WAN, the service levels and bandwidth may vary, and the shared connections do not guarantee security.

WAN Topologies (7.1.3)

A physical topology describes the physical network infrastructure used by data when it is traveling from a source to a destination. The physical WAN topology used in WANs is complex and, for the most part, unknown to users. Consider a user in New York establishing a video conference call with a user in Tokyo, Japan. Other than the user’s internet connection in New York, it would not be feasible to identify all the actual physical connections needed to support the video call.

A WAN topology is described using a logical topology, which includes the virtual connection between the source and destination. For example, a video conference call between a user in New York and a user in Japan would be a logical point-to-point connection.

WANs are implemented using the following logical topology designs:

• Point-to-point topology

• Hub-and-spoke topology

• Dual-homed topology

• Fully meshed topology

• Partially meshed topology

Point-to-Point Topology

A point-to-point topology, as shown in Figure 7-2, uses a point-to-point circuit between two endpoints.

Point-to-point links often involve dedicated, leased-line connections from the corporate edge point to the provider networks. A point-to-point connection involves a Layer 2 transport service through the service provider network. Packets sent from one site are delivered to the other site and vice versa. A point-to-point connection is transparent to the customer network. It seems as if there is a direct physical link between two endpoints.

Point-to-point topology can become expensive if many point-to-point connections are required.

Hub-and-Spoke Topology

A hub-and-spoke topology enables a single interface on the hub router to be shared by all spoke circuits. Spoke routers can be interconnected through the hub router using virtual circuits and routed subinterfaces. Figure 7-3 displays a sample hub-and-spoke topology consisting of three spoke routers connecting to a hub router across a WAN cloud.

A hub-and-spoke topology is a single-homed topology. There is only one hub router, and all communication must go through it. Therefore, spoke routers can only communicate with each other through the hub router. Consequently, the hub router represents a single point of failure. If it fails, inter-spoke communication also fails.

Dual-homed Topology

A dual-homed topology provides redundancy. Figure 7-4 shows such a topology, with two hub routers dual-homed and redundantly attached to three spoke routers across a WAN cloud.

The advantage of dual-homed topologies is that they offer enhanced network redundancy, load balancing, distributed computing and processing, and the ability to implement backup service provider connections.

The disadvantage is that they are more expensive to implement than single-homed topologies. This is because they require additional networking hardware, such as additional routers and switches. Dual-homed topologies are also more difficult to implement because they require additional, and more complex, configurations.

Fully Meshed Topology

A fully meshed topology uses multiple virtual circuits to connect all sites, as shown in Figure 7-5.

This is the most fault-tolerant topology of the five shown in this chapter. For instance, if site B lost connectivity to site A, it could send the data through either site C or site D.

Partially Meshed Topology

A partially meshed topology connects many but not all sites. For example, in Figure 7-6, sites A, B, and C are fully meshed, but site D must connect to site A to reach sites B and C.

Carrier Connections (7.1.4)

An important aspect of WAN design is how an organization connects to the internet. An organization usually signs a service level agreement (SLA) with a service provider. The SLA outlines the expected services related to the reliability and availability of the connection. The service provider may or may not be the actual carrier. A carrier owns and maintains the physical connection and equipment between the provider and the customer. Typically, an organization chooses either a single-carrier or dual-carrier WAN connection.

Single-Carrier WAN Connection

With a single-carrier connection, an organization connects to only one service provider, as shown in Figure 7-7. An SLA is negotiated between the organization and the service provider. The disadvantage of this design is that the carrier connection and service provider are both single points of failure. Connectivity to the internet would be lost if the carrier link failed or if the provider router failed.

Dual-Carrier WAN Connection

A dual-carrier connection provides redundancy and increases network availability, as shown in Figure 7-8. The organization negotiates separate SLAs with two different service providers. The organization should ensure that the two providers each use a different carrier. Although a dual-carrier connection is more expensive to implement than a single-carrier connection, the second connection can be used for redundancy (that is, as a backup link). It can also be used to improve network performance and load balance internet traffic.

Evolving Networks (7.1.5)

A company’s network requirements can change dramatically as the company grows over time. Distributing employees saves costs in many ways, but it puts increased demands on the network. A network must meet the day-to-day operational needs of the business, and it also must be able to adapt and grow as the company changes. Network designers and administrators meet these challenges by carefully choosing network technologies, protocols, and service providers. They must also optimize their networks by using a variety of network design techniques and architectures.

To illustrate differences between networks of various sizes, in this chapter we use a fictitious company called SPAN Engineering and watch as it grows from a small local business into a global enterprise. SPAN Engineering, an environmental consulting firm, has developed a special process for converting household waste into electricity and is developing a small pilot project for a municipal government in its local area.

Small Network

SPAN Engineering initially consisted of 15 employees working in a small office, as shown in Figure 7-9.

They used a single LAN connected to a wireless router for sharing data and peripherals. The connection to the internet was through a common broadband service called digital subscriber line (DSL), supplied by the local telephone service provider. To support the company IT requirements, SPAN contracted services from the DSL provider.

Campus Network

Within a few years, SPAN Engineering grew to require several floors of a building, as shown in Figure 7-10.

The company now required a campus-area network (CAN). A CAN interconnects several LANs within a limited geographic area. Multiple LANs are required to segment the various departments that are connecting to multiple switches in a campus network environment.

The network includes dedicated servers for email, data transfer, file storage, and web-based productivity tools and applications. A firewall secures internet access to corporate users. The business now requires in-house IT staff to support and maintain the network.

Branch Network

A few years later, SPAN Engineering expanded and added a branch site in the city and remote and regional sites in other cities, as shown in Figure 7-11.

The company now required a metropolitan-area network (MAN) to interconnect sites within the city. A MAN is larger than a LAN but smaller than a WAN.

To connect to the central office, branch offices in nearby cities used private dedicated lines through their local service provider. Offices in other cities and countries require the services of a WAN or may use internet services to connect distant locations. However, the internet introduces security and privacy issues that the IT team must address.

Distributed Network

SPAN Engineering has now been in business for 20 years and has grown to having thousands of employees distributed in offices worldwide, as shown in Figure 7-12.

To reduce network costs, SPAN encouraged teleworking and virtual teams using web-based applications, including web conferencing, e-learning, and online collaboration tools to increase productivity and reduce costs. Site-to-site and remote-access virtual private networks (VPNs) enable the company to use the internet to connect easily and securely with employees and facilities around the world.

Check Your Understanding—Purpose of WANs (7.1.6)

Refer to the online course to complete this activity.

WAN Standards (7.2.1)

Modern WAN standards are defined and managed by a number of recognized authorities, including the following:

• TIA/EIA, which is a combination of the Telecommunications Industry Association (TIA) and Electronic Industries Alliance (EIA)

• International Organization for Standardization (ISO)

• Institute of Electrical and Electronics Engineers (IEEE)

WANs in the OSI Model (7.2.2)

Most WAN standards focus on the physical layer (OSI Layer 1) and the data link layer (OSI Layer 2), as shown in Figure 7-13.

Common WAN Terminology (7.2.3)

The WAN physical layer describes the physical connections between the company network and the service provider network.

Specific terms are used to describe WAN connections between the subscriber that is, the company/client) and the WAN service provider, as shown in Figure 7-14.

Table 7-2 explains the terms shown in Figure 7-14, as well as some additional WAN-related terms.

WAN Devices (7.2.4)

Many types of devices are specific to WAN environments. However, the end-to-end data path over a WAN is usually from the source DTE to the DCE, then to the WAN cloud, then to the DCE, and finally to the destination DTE, as shown in Figure 7-15.

Table 7-3 describes the WAN devices shown in Figure 7-15.

Serial Communication (7.2.5)

Almost all network communications occur using serial communication delivery. Serial communication transmits bits sequentially over a single channel. In contrast, parallel communications simultaneously transmit several bits using multiple wires.

Figure 7-16 demonstrates the difference between serial and parallel communications.

Although a parallel connection theoretically transfers data eight times faster than a serial connection, it is prone to synchronization problems. As the cable length increases, the synchronization timing between multiple channels becomes more sensitive to distance. For this reason, parallel communication is limited to very short distances only; for example, copper media is limited to less than 8 meters (that is, 26 feet).

Parallel communication is not a viable WAN communication method because of its length restriction. It is, however, a viable solution in data centers, where distances between servers and switches are relatively short. For instance, the Cisco Nexus switches in data centers support parallel optics solutions to transfer more data signals and achieve higher speeds (such as 40 Gbps and 100 Gbps).

Circuit-Switched Communication (7.2.6)

Network communication can be implemented using circuit-switched communication. A circuit-switched network establishes a dedicated circuit (or channel) between endpoints before the users can communicate. Specifically, circuit switching dynamically establishes a dedicated virtual connection through the service provider network before voice or data communication can start, as shown in Figure 7-17.

For example, when a user makes a telephone call using a landline, the number called is used by the provider equipment to create a dedicated circuit from the caller to the called party.

During transmission over a circuit-switched network, all communication uses the same path. The entire fixed capacity allocated to the circuit is available for the duration of the connection, regardless of whether there is information to transmit. This can lead to inefficiencies in circuit usage. For this reason, circuit switching is generally not suited for data communication.

The two most common types of circuit-switched WAN technologies are the public switched telephone network (PSTN) and the legacy Integrated Services Digital Network (ISDN).

Packet-Switched Communications (7.2.7)

Network communication is most commonly implemented using packet-switched communication. In contrast to circuit switching, packet switching segments traffic data into packets that are routed over a shared network. Packet-switched networks do not require a circuit to be established, and they allow many pairs of nodes to communicate over the same channel, as shown in Figure 7-18.

Packet switching is much less expensive and more flexible than circuit switching. Although susceptible to delays (latency) and variability of delay (jitter), modern packet-switching technology allows satisfactory transport of voice and video communications on these networks.

Examples of packet-switched WAN technologies are Ethernet WAN (Metro Ethernet) and Multiprotocol Label Switching (MPLS), as well as the legacy technologies Frame Relay and Asynchronous Transfer Mode (ATM).

SDH, SONET, and DWDM (7.2.8)

Service provider networks use fiber-optic infrastructures to transport user data between destinations. Fiber-optic cable is far superior to copper cable for long-distance transmissions due to its much lower attenuation and interference.

There are two optical fiber OSI Layer 1 standards available to service providers:

• SDH: Synchronous Digital Hierarchy (SDH) is a global standard for transporting data over fiber-optic cable.

• SONET: Synchronous Optical Networking (SONET) is the North American standard that provides the same services as SDH.

These two standards are essentially the same, and they are therefore often listed as SONET/SDH.

SONET and SDH define how to transfer multiple data, voice, and video communications over optical fiber using lasers or light-emitting diodes (LEDs) over great distances. Both standards are used on the ring network topology, which contains the redundant fiber paths that allow traffic to flow in both directions.

Dense wavelength-division multiplexing (DWDM) is a newer technology that increases the data-carrying capacity of SDH and SONET by simultaneously sending multiple streams of data (multiplexing) using different wavelengths of light, as shown in Figure 7-19.

DWDM has the following features:

• It supports SONET and SDH standards.

• It can multiplex more than 80 different channels of data (that is, wavelengths) onto a single fiber.

• Each channel is capable of carrying a 10 Gbps multiplexed signal.

• It assigns incoming optical signals to specific wavelengths of light (that is, frequencies).

Traditional WAN Connectivity Options (7.3.1)

When LANs appeared in the 1980s, organizations began to see the need to interconnect with other locations. To do so, they needed their networks to connect to the local loop of a service provider. This was accomplished by using dedicated lines or by using switched services from a service provider.

Figure 7-20 summarizes the traditional WAN connectivity options.

Common WAN Terminology (7.3.2)

When permanent dedicated connections were required, a point-to-point link using copper media was used to provide a preestablished WAN communications path from the customer premises to the provider network. Point-to-point lines could be leased from a service provider. These lines are called leased lines because the organization pays a monthly lease fee to a service provider to use the lines.

Leased lines have existed since the early 1950s. They are referred to by different names, such as leased circuits, serial links, serial lines, point-to-point links, and T1/E1 or T3/E3 lines.

Leased lines are available in different fixed capacities and are generally priced based on the bandwidth required and the distance between the two connected points.

There are two systems used to define the digital capacity of a copper media serial link:

• T-carrier: Used in North America, T-carrier provides T1 links supporting bandwidth up to 1.544 Mbps and T3 links supporting bandwidth up to 43.7 Mbps.

• E-carrier: Used in Europe, E-carrier provides E1 links supporting bandwidth up to 2.048 Mbps and E3 links supporting bandwidth up to 34.368 Mbps.

Table 7-4 summarizes the advantages and disadvantages of leased lines.

Circuit-Switched Options (7.3.3)

Circuit-switched connections are provided by public switched telephone network (PSTN) carriers. The local loop connecting the CPE to the CO is copper media. There are two traditional circuit-switched options: the PSTN and ISDN.

Packet-Switched Options (7.3.4)

Packet switching segments data into packets that are routed over a shared network. Circuit-switched networks require a dedicated circuit to be established. In contrast, packet-switching networks allow many pairs of nodes to communicate over the same channel.

There are two traditional (legacy) packet-switched connectivity options: Frame Relay and ATM.

Modern WANs (7.4.1)

Modern WANs have more connectivity options than traditional WANs. Enterprises now require faster and more flexible WAN connectivity options. Traditional WAN connectivity options have rapidly declined in use because they are either no longer available, too expensive, or have limited bandwidth.

Figure 7-21 displays the local loop connections most likely to be encountered today.

Modern WAN Connectivity Options (7.4.2)

New technologies are continually emerging. Figure 7-22 summarizes the modern WAN connectivity options.

Ethernet WAN (7.4.3)

Ethernet was originally developed as a LAN access technology and was not suitable as a WAN access technology due primarily to the limited distance provided by copper media.

However, newer Ethernet standards using fiber-optic cables have made Ethernet a reasonable WAN access option. For instance, the IEEE 1000BASE-LX standard supports fiber-optic cable lengths of 5 km, and the IEEE 1000BASE-ZX standard supports cable lengths up to 70 km.

Service providers now offer Ethernet WAN service using fiber-optic cabling. The Ethernet WAN service can go by many names, including the following:

• Metropolitan Ethernet (MetroE)

• Ethernet over MPLS (EoMPLS)

• Virtual Private LAN Service (VPLS)

Figure 7-23 shows a simple Metro Ethernet topology example.

The following are several benefits of an Ethernet WAN:

• Reduced expenses and administration: An Ethernet WAN provides a switched, high-bandwidth Layer 2 network capable of managing data, voice, and video all on the same infrastructure. This increases bandwidth and eliminates expensive conversions to other WAN technologies. The technology enables businesses to inexpensively connect numerous sites in a metropolitan area to each other and to the internet.

• Easy integration with existing networks: An Ethernet WAN connects easily to existing Ethernet LANs, reducing installation costs and time.

• Enhanced business productivity: An Ethernet WAN enables businesses to take advantage of productivity-enhancing IP applications that are difficult to implement on TDM or Frame Relay networks, such as hosted IP communications, VoIP, and streaming and broadcast video.

MPLS (7.4.4)

Multiprotocol Label Switching (MPLS) is a high-performance service provider WAN routing technology to interconnect clients without regard to access method or payload. MPLS supports a variety of client access methods (for example, Ethernet, DSL, cable, Frame Relay). MPLS can encapsulate all types of protocols, including IPv4 and IPv6 traffic.

The sample topology in Figure 7-24 is a simple MPLS-enabled network.

An MPLS router can be a customer edge (CE) router, a provider edge (PE) router, or an internal provider (P) router. Notice that MPLS supports a variety of client access connections.

MPLS routers are label switched routers (LSRs). This means that they attach labels to packets that are then used by other MPLS routers to forward traffic. When traffic is leaving the CE, the MPLS PE router adds a short fixed-length label in between the frame header and the packet header. MPLS P routers use the label to determine the next hop of the packet. The label is removed by the egress PE router when the packet leaves the MPLS network.

MPLS also provides services for QoS support, traffic engineering, redundancy, and VPNs.

Internet-Based Connectivity Options (7.5.1)

Modern WAN connectivity options do not end with Ethernet WAN and MPLS. Today, there are a host of internet-based wired and wireless options from which to choose. Internet-based broadband connectivity is an alternative to using dedicated WAN options.

Figure 7-25 lists the internet-based connectivity options.

Internet-based connectivity can be divided into wired and wireless options.

DSL Technology (7.5.2)

Digital subscriber line (DSL) is a high-speed, always-on connection technology that uses existing twisted-pair telephone lines to provide IP services to users. DSL is a popular choice for home users and for enterprise IT departments to support teleworkers.

Figure 7-26 shows a representation of bandwidth space allocation on a copper wire for asymmetric DSL (ADSL).

The area of the figure labeled POTS (plain old telephone system) identifies the frequency range used by the voice-grade telephone service. The area labeled ADSL represents the frequency space used by the upstream and downstream DSL signals. The area that encompasses both the POTS area and the ADSL area represents the entire frequency range supported by the copper wire pair.

There are several xDSL varieties offering different upload and download transmission rates. However, all forms of DSL are categorized as either asymmetric DSL (ADSL) or symmetric DSL (SDSL). ADSL and ADSL2+ provide higher downstream bandwidth to the user than upload bandwidth. SDSL provides the same capacity in both directions.

The transfer rates are also dependent on the actual length of the local loop and the type and condition of the cabling. For example, an ADSL loop must be less than 5.46 km (3.39 miles) for guaranteed signal quality.

Security risks are incurred with DSL technology but can be mediated with security measures such as VPNs.

DSL Connections (7.5.3)

Service providers deploy DSL connections in the local loop. As shown in Figure 7-27, a connection is set up between the DSL modem and the DSL access multiplexer (DSLAM).

A DSL modem converts the Ethernet signals from the teleworker device to a DSL signal, which is transmitted to a DSLAM at the provider location.

A DSLAM is the device located at the CO of the provider and concentrates connections from multiple DSL subscribers. A DSLAM is often built into an aggregation router.

The advantage that DSL has over cable technology is that DSL is not a shared medium. Each user has a separate direct connection to the DSLAM. Adding users does not impede performance, unless the DSLAM internet connection to the ISP or to the internet becomes saturated.

DSL and PPP (7.5.4)

Point-to-Point Protocol (PPP) is a Layer 2 protocol that was commonly used by telephone service providers to establish router-to-router and host-to-network connections over dialup and ISDN access networks.

ISPs still use PPP as the Layer 2 protocol for broadband DSL connections for several reasons:

• PPP can be used to authenticate the subscriber.

• PPP can assign a public IPv4 address and/or IPv6 prefix to the subscriber.

• PPP provides link-quality management features.

In PPP over Ethernet (PPPoE), a DSL modem has a DSL interface to connect to the DSL network and an Ethernet interface to connect to the client device. However, Ethernet links do not natively support PPP.

Host with PPPoE Client

In the example in Figure 7-28, the host runs a PPPoE client to obtain a public IPv4 address and/or IPv6 prefix from a PPPoE server located at the provider site. The PPPoE client software communicates with the DSL modem using PPPoE, and the modem communicates with the ISP using PPP. In this topology, only one client can use the connection. Also, notice that there is no router to protect the inside network.

Router PPPoE Client

Another solution is to configure a router to be a PPPoE client, as shown in Figure 7-29. The router is the PPPoE client and obtains its configuration from the provider. The clients communicate with the router using only Ethernet and are unaware of the DSL connection. In this topology, multiple clients can share the DSL connection.

Cable Technology (7.5.5)

Cable technology is a high-speed always-on connection technology that uses a coaxial cable from the cable company to provide IP services to users. Like DSL, cable technology is a popular choice for home users and for enterprise IT departments to support remote workers.

Modern cable systems offer customers advanced telecommunications services, including high-speed internet access, digital cable television, and residential telephone service.

Data over Cable Service Interface Specification (DOCSIS) is the international standard for adding high-bandwidth data to an existing cable system.

Cable operators deploy hybrid fiber-coaxial (HFC) networks to enable high-speed transmission of data to cable modems. The cable system uses a coaxial cable to carry radio frequency (RF) signals to the end user.

HFC uses fiber-optic and coaxial cable in different portions of the network. For example, the connection between the cable modem and optical node is coaxial cable, as shown in Figure 7-30.

The optical node performs optical-to-RF signal conversion. Specifically, it converts RF signals to light pulses over fiber-optic cable. The fiber media enables the signals to travel over long distances to the provider headend, where a cable modem termination system (CMTS) is located. The headend contains the databases needed to provide internet access, and the CMTS is responsible for communicating with the cable modems.

All the local subscribers share the same cable bandwidth. As more users join the service, available bandwidth may drop below the expected rate.

Optical Fiber (7.5.6)

Many municipalities, cities, and providers install fiber-optic cable to the user location. This is commonly referred to as fiber-to-the-x (FTTx) and includes the following:

• Fiber-to-the-home (FTTH): Fiber reaches the boundary of the residence. Passive optical networks and point-to-point Ethernet are architectures that can deliver cable TV, internet, and phone services over FTTH networks directly from the service provider central office.

• Fiber-to-the-building (FTTB): Fiber reaches the boundary of the building, such as the basement in a multi-dwelling unit, with the final connection to the individual living space being made via alternative means, like curb or pole technologies.

• Fiber-to-the-node/neighborhood (FTTN): Optical cabling reaches an optical node that converts optical signals to a format acceptable for twisted pair or coaxial cable to the premises.

FTTx can deliver the highest bandwidth of all broadband options.

Wireless Internet-Based Broadband (7.5.7)

Wireless technology uses the unlicensed radio spectrum to send and receive data. The unlicensed spectrum is accessible to anyone who has a wireless router and wireless technology in the device he or she is using.

Until recently, one limitation of wireless access has been the need to be within the local transmission range (typically less than 100 feet) of a wireless router or a wireless modem with a wired connection to the internet.

VPN Technology (7.5.8)

Security risks are incurred when a teleworker or a remote office worker uses a broadband service to access the corporate WAN over the internet.

To address security concerns, broadband services provide virtual private network (VPN) connections to a network device that accepts VPN connections. The network device is typically located at the corporate site.

A VPN is an encrypted connection between private networks over a public network, such as the internet. Instead of using a dedicated Layer 2 connection, such as a leased line, a VPN uses virtual connections called VPN tunnels. VPN tunnels are routed through the internet from the private network of the company to the remote site or employee host.

The following are several benefits to using VPN:

• Cost savings: VPNs enable organizations to use the global internet to connect remote offices and to connect remote users to the main corporate site. This eliminates the need for expensive, dedicated WAN links and modem banks.

• Security: VPNs provide the highest level of security by using advanced encryption and authentication protocols that protect data from unauthorized access.

• Scalability: Because VPNs use the internet infrastructure within ISPs and devices, it is easy to add new users. Corporations can add large amounts of capacity without adding significant infrastructure.

• Compatibility with broadband technology: VPN technology is supported by broadband service providers such as DSL and cable. VPNs allow mobile workers and telecommuters to take advantage of their home high-speed internet service to access their corporate networks. Business-grade, high-speed broadband connections can also provide a cost-effective solution for connecting remote offices.

VPNs are commonly implemented as the following:

• Site-to-site VPN: VPN settings are configured on routers. Clients are unaware that their data is being encrypted.

• Remote access VPN: The user initiates remote-access connection (such as by using HTTPS in a browser to connect to the bank). Alternatively, the user can run VPN client software on his or her host to connect to and authenticate with the destination device.

ISP Connectivity Options (7.5.9)

This section explains the different ways an organization can connect to an ISP. The choice depends on the needs and budget of the organization.

Single-Homed

Single-homed ISP connectivity is used by the organization when internet access is not crucial to the operation. As shown in Figure 7-31, the client connects to the ISP using one link. This topology provides no redundancy, and it is the least expensive solution of the four described in this section.

Dual-Homed

Dual-homed ISP connectivity is used by an organization when internet access is somewhat crucial to the operation. As shown in Figure 7-32, the client connects to the same ISP using two links. The topology provides both redundancy and load balancing. If one link fails, the other link can carry the traffic. If both links are operational, traffic can be load balanced over them. However, the organization loses internet connectivity if the ISP experiences an outage.

Multihomed

Multihomed ISP connectivity is used by an organization when internet access is crucial to the operation. The client connects to two different ISPs, as shown in Figure 7-33. This design provides increased redundancy and enables load balancing, but it can be expensive.

Dual-Multihomed

Dual-multihomed ISP is the most resilient topology of the four described in this section. The client connects with redundant links to multiple ISPs, as shown in Figure 7-34. This topology provides the most redundancy possible, and it is the most expensive option of the four.

Broadband Solution Comparison (7.5.10)

Each broadband solution has advantages and disadvantages. The ideal solution is to have a fiber-optic cable directly connected to a client network. Some locations have only one option, such as cable or DSL. Some locations have only broadband wireless options for internet connectivity.

If there are multiple broadband solutions available, a cost-versus-benefit analysis should be performed to determine the best solution.

Some factors to consider include the following:

• Cable: Bandwidth is shared by many users. Therefore, upstream data rates are often slow during high-usage hours in areas with oversubscription.

• DSL: DSL has limited bandwidth and is distance sensitive (in relation to the ISP central office). The upload rate is proportionally lower than the download rate.

• Fiber-to-the-home: This option requires fiber installation directly to the home.

• Cellular/mobile: With this option, coverage is often an issue, even within a small office or home office where bandwidth is relatively limited.

• Municipal Wi-Fi: Most municipalities do not have a mesh Wi-Fi network deployed. If municipal Wi-Fi is available and in range, however, it is a viable option.

• Satellite: This option is expensive and provides limited capacity per subscriber. It is typically used when no other option is available.

Purpose of WANs

A wide-area network (WAN) is required to connect beyond the boundaries of a LAN. A WAN is a telecommunications network that spans a relatively large geographic area. A WAN operates beyond the geographic scope of a LAN. A private WAN is a connection that is dedicated to a single customer. A public WAN connection is typically provided by an ISP or a telecommunications service provider using the internet. A WAN topology is described using a logical topology. WANs are implemented using the following logical topologies: point-to-point, hub-and-spoke, dual-homed, fully meshed, and partially meshed. With a single-carrier connection, an organization connects to only one service provider. A dual-carrier connection provides redundancy and increases network availability. The organization negotiates separate SLAs with two different service providers. Network requirements of a company can change dramatically as the company grows over time. Distributing employees saves costs in many ways, but it puts increased demands on the network. Small companies may use a single LAN connected to a wireless router to share data and peripherals. Connection to the internet is through a broadband service provider. A slightly larger company may use a campus-area network (CAN), which interconnects several LANs within a limited geographic area. An even larger company may require a metropolitan-area network (MAN) to interconnect sites within the city. A MAN is larger than a LAN but smaller than a WAN. A global company may require teleworking and virtual teams using web-based applications, including web conferencing, e-learning, and online collaboration tools. Site-to-site and remote-access virtual private networks (VPNs) enable a company to use the internet to securely connect with employees and facilities around the world.

WAN Operations

Modern WAN standards are defined and managed by a number of recognized authorities, including TIA/EIA, ISO, and IEEE. Most WAN standards focus on the physical layer (OSI Layer 1) and the data link layer (OSI Layer 2). Layer 1 protocols describe the electrical, mechanical, and operational components needed to transmit bits over a WAN. Layer 1 optical fiber protocol standards include SDH, SONET, and DWDM. Layer 2 protocols define how data is encapsulated into a frame. Layer 2 protocols include broadband, wireless, Ethernet WAN, MPLS, PPP, and HDLC. The WAN physical layer describes the physical connections between a company network and a service provider network. There are specific terms used to describe WAN connections between the subscriber (that is, the company/client) and a WAN service provider, including DTE, DCE, CPE, POP, demarcation point, local loop, CO, toll network, backhaul network, and backbone network. The end-to-end data path over a WAN is usually from source DTE to the DCE, then to the WAN cloud, then to the DCE, and finally to the destination DTE. Devices used in this path include voiceband modems, DSL and cable modems, CSUs/DSUs, optical converters, wireless routers or access points, and other WAN core devices. Serial communication transmits bits sequentially over a single channel. In contrast, parallel communications simultaneously transmit several bits using multiple wires. A circuit-switched network establishes a dedicated circuit (or channel) between endpoints before the users can communicate. During transmission over a circuit-switched network, all communication uses the same path. The two most common types of circuit-switched WAN technologies are PSTN and ISDN. Packet-switching segments traffic data into packets that are routed over a shared network. Common types of packet-switched WAN technologies are Ethernet WAN and MPLS. There are two optical fiber OSI Layer 1 standards. SDH/SONET define how to transfer multiple data, voice, and video communications over optical fiber using lasers or LEDs over great distances. Both standards are used on the ring network topology, which contains redundant fiber paths allowing traffic to flow in both directions. DWDM is a newer technology that increases the data-carrying capacity of SDH and SONET by simultaneously sending multiple streams of data (multiplexing) using different wavelengths of light.

Traditional WAN Connectivity

In the 1980s, organizations started to see the need to interconnect their LANs with other locations. They needed their networks to connect to the local loop of a service provider by using dedicated lines or by using switched services from a service provider. When permanent dedicated connections were required, a point-to-point link using copper media was used to provide a preestablished WAN communications path from the customer premises to the provider network. Dedicated leased lines were T1/E1 or T3/E3 lines. Circuit-switched connections were provided by PSTN carriers. The local loop connecting the CPE to the CO was copper media. ISDN is a circuit-switching technology that enables the PSTN local loop to carry digital signals. It provided higher-capacity switched connections than dialup access. Packet switching segments data into packets that are routed over a shared network. Packet-switching networks allow many pairs of nodes to communicate over the same channel. Frame Relay is a simple Layer 2 NBMA WAN technology used to interconnect enterprise LANs. ATM technology is capable of transferring voice, video, and data through private and public networks. It is built on a cell-based architecture rather than on a frame-based architecture.

Modern WAN Connectivity

Modern WAN connectivity options include dedicated broadband, Ethernet WAN, and MPLS (packet switched), along with various wired and wireless versions of internet-based broadband. Service providers now offer Ethernet WAN service using fiber-optic cabling. Ethernet WAN reduces expenses and administration, is easily integrated with existing networks, and enhances business productivity. MPLS is a high-performance service provider WAN routing technology for interconnecting clients. MPLS supports a variety of client access methods (such as Ethernet, DSL, cable, and Frame Relay). MPLS can encapsulate all types of protocols, including IPv4 or IPv6 traffic.

Internet-Based Connectivity

Internet-based broadband connectivity is an alternative to using dedicated WAN options. There are wired and wireless versions of broadband VPN. Wired options use permanent cabling (such as copper or fiber) to provide consistent bandwidth and reduce error rates and latency. Examples of wired broadband connectivity are digital subscriber line (DSL), cable connections, and optical fiber networks. Examples of wireless broadband include cellular 3G/4G/5G and satellite internet services. DSL is a high-speed, always-on connection technology that uses existing twisted-pair telephone lines to provide IP services to users. All forms of DSL are categorized as either ADSL or SDSL. A DSL modem converts Ethernet signals from a teleworker’s device to a DSL signal, which is transmitted to a DSLAM at the provider location. The advantage that DSL has over cable technology is that DSL is not a shared medium. ISPs still use PPP as the Layer 2 protocol for broadband DSL connections. A DSL modem has a DSL interface to connect to the DSL network and an Ethernet interface to connect to the client device. Ethernet links do not natively support PPP. Cable technology is a high-speed always-on connection technology that uses a cable company coaxial cable to provide IP services to users. Cable operators deploy hybrid fiber-coaxial (HFC) networks to enable high-speed transmission of data to cable modems. The cable system uses a coaxial cable to carry radio frequency (RF) signals to the end user. Many municipalities, cities, and providers install fiber-optic cable to the user location. This is commonly referred to as fiber-to-the-x (FTTx), such as FTTH, FTTB, and FTTN.

Wireless technology uses the unlicensed radio spectrum to send and receive data. The unlicensed spectrum is accessible to anyone who has a wireless router and wireless technology in the device he or she is using. Until recently, one limitation of wireless access has been the need to be within the local transmission range (typically less than 100 feet) of a wireless router or a wireless modem that has a wired connection to the internet. Newer developments in wireless technology include municipal Wi-Fi, cellular, satellite internet, and WiMAX. To address security concerns, broadband services provide capabilities for using virtual private network (VPN) connections to a network device that accepts VPN connections, which is typically located at the corporate site. A VPN is an encrypted connection between private networks over a public network, such as the internet. Instead of using a dedicated Layer 2 connection, such as a leased line, a VPN uses virtual connections called VPN tunnels. VPN tunnels are routed through the internet from the private network of the company to the remote site or employee host. Common VPN implementations include site-to-site and remote-access VPNs. ISP connectivity options include single-homed, dual-homed, multihomed, and dual-multihomed. Cable, DSL, fiber-to-the-home, cellular/mobile, municipal Wi-Fi, and satellite internet all have advantages and disadvantages. Perform a cost-versus-benefit analysis before choosing an internet-based connectivity solution.