Wide Area Network What is WAN (Wide Area Network)?
A wide area network (WAN) can connect multiple computers together across a large geographical area, often spanning multiple cities or even countries. Typically, organizations use private WAN links to connect their branch offices to their headquarters or their corporate data center. In most cases, organizations don’t build WAN connections themselves, and instead rent leased lines to service providers. Technologies such as SD-WAN and Multiprotocol Label Switching (MPLS) are often used in WAN connections. Other technologies were used in the past such as X.25, Frame Relay and Asynchronous Transfer Mode (ATM).
- WAN explained
- How does WAN work?
- History of WAN
- Traditional WAN vs SD-WAN
WAN explained
WAN interconnects multiple Local Area Networks (LAN) over large geographical areas spanning cities, states and even continents using routers installed at each end of the network.
There are two types of WAN: private and public. A private WAN is a network owned and operated by a single organization. It is used to connect devices that are in different parts of the organization, such as different branch offices. A private WAN is typically built using leased lines or dedicated circuits, which provide a high level of security and reliability.
A public WAN, on the other hand, is a network used to connect devices belonging to different organizations. The most common example of a public WAN is the Internet.
How does WAN work?
WAN relies on a variety of technologies, including optical fiber, satellites, microwave links, and circuit-switched telephone lines. As technology evolves, the speed on these links has increased dramatically. Speed over fiber optic links can typically reach speeds of up to 100 Gbps and even higher.
Data is transmitted in packets over the WAN. A packet is a small unit of data that contains information about the source, destination, and the data itself. When a device wants to send data over a WAN, it splits the data into packets and sends them over the network. The packets are then recombined at destination.
Additionally, due to the distance between the source and destination, WAN is often prone to latency effects impacting network performance. To overcome these effects, WAN transmission can be improved with techniques such as WAN optimization including TCP protocol acceleration, data deduplication or data compression.
History of WAN
The history of WAN has been marked by various technological improvements that successively increased transmission rates. In the 1980’s, network speeds were counted in Kbit/s. Today, internet connections can reach speeds as high as 100 Gbit/s.
Here are some of the main technologies used to connect wide area networks.
- X.25: In the 1970’s, the International Telegraph and Telephone Consultative Committee (CCITT, now ITU-T) developed the X.25 protocol. It is the oldest packet-switched data communication protocol (method to group data into packets) and was used until 2015. It uses a point-to-point architecture that served to connect remote terminals to mainframes. It operated over analog channels leased from telephone companies.
- Frame Relay: In the 1980’s, Frame Relay became an alternative to X.25 providing increasing speed. By improving voice and video performance, it was largely adopted by enterprises in the US while X.25 remained the standard in Europe. Frame Relay transmits data in variable-size units called “frames” over a virtual circuit-based connection. It does not perform any error correction such as retransmission of data, and leaves error checks to terminals. If an error is detected, the packet is simply dropped.
- ATM: Asynchronous Transfer Mode or ATM was developed in the late 1980’s and early 1990’s. It differs from Frame Relay by transferring fixed size cells (53 bytes). Also, ATM provided error correction unlike Frame Relay and was faster (up to 622 Mbps compared to 45 Mbps for Frame Relay). However, ATM didn’t get the expected success due to the better price performance of internet-protocol based products as well as the size of cells (53 bytes) that was not efficient.
- MPLS: Multiprotocol Label Switching (MPLS) was developed in the late 1990’s as a more flexible and scalable alternative to ATM. MPLS routes packets based on labels and not IP addresses. It can be used with any networking protocol, including Ethernet, ATM, and Frame Relay. Even though MPLS provides scalability and performance, it fails shorts to support modern cloud architectures as SaaS traffic needs to be backhauled to the data center for security inspection, impacting application performance.
- SD-WAN: Developed in the 2010’s, it can combine heterogeneous links, including MPLS, broadband internet and 5G through network virtualization, providing redundancy and improved performance. It allows organizations to reduce MPLS dependency by leveraging low-cost internet connections. With SD-WAN, branch offices can also better manage SaaS traffic by avoiding sending cloud traffic back to the data center. SD-WAN is also part of SASE (Secure Access Service Edge), ensuring secure access to cloud applications, from anywhere and any device.
Traditional WAN vs SD-WAN
While both WAN and Software-Defined WAN (SD-WAN) are used to connect devices over a large area, there are significant differences between the two. WAN is a traditional networking technology that uses physical connections such as leased lines and satellite links to connect devices. In contrast, SD-WAN is a newer technology that uses software to manage and optimize the flow of data over the network.
SD-WAN can combine multiple links including MPLS, broadband internet and 5G, increasing network bandwidth and performance. SD-WAN is flexible and can operate over any links. Not only does SD-WAN allow organizations to spin up new branches quickly but also offers improved performance for high demanding applications such as voice and video applications. Advanced SD-WAN solutions indeed include optimization techniques to overcome the adverse effects of packet loss and jitter often found on broadband internet links. For example, Forward Error Correction (FEC) is able to rebuild packets at destination using parity packets offering private-line-like performance over broadband internet links. SD-WAN also provides redundancy for critical applications by combining multiple links based on business requirements, but also by using some specific links as failover eliminating potential brownouts or blackouts.
Advanced SD-WAN solutions also include in a single platform other functionality such as WAN Optimization or next-generation firewalls, enabling organizations to drastically reduce hardware footprint in branch offices.
SD-WAN supports modern cloud architectures, now largely adopted by organizations. It intelligently steers SaaS traffic to the cloud without backhauling traffic to the data center. Trusted applications are sent directly to the cloud, other traffic is sent to SSE services (Security Service Edge) in a SASE architecture. Advanced SD-WAN can even be deployed in cloud providers like AWS, Microsoft Azure and Google Cloud improving application performance and security.