Introduction to Ethernet
An introduction to ethernet and how it works over a SOHO LAN
Ethernet
Ethernet is by far the most widely used form of networking. Within a Small Office/ Home Office (SOHO) Local Area Network (LAN), ethernet is king due to its reliability, reasonable costs, and super quick speeds. This type of environment is one you might expect in your own home or at your favorite café.
Equipment Used to Set Up a SOHO LAN
Typically, within a small office or home office, you or your business would start with an ordinary router. Today, routers have been designed to serve multiple functions across multiple layers of the network. The functions include: routing (L3), providing a wireless access point (L2), modem functions (L2), firewalls (L4), and even switch capabilities (L3).
Check out my previous article on the OSI Model to learn about the layers I’m referring to here:
https://gabertan-colton.medium.com/an-overview-the-osi-model-813e817108fb
A basic router:
Routers are what connect your internal network to other networks across the world via the internet. They typically don’t require much configuration in terms of its switching and routing preferences; however, you may need to configure features such as WPA2 for security or the frequency that you wish the wireless standard to operate on.
Because routers have become such useful all-in-one devices for small networks such as a SOHO LAN, all that’s left is the ethernet cabling itself. The standard today would be to use copper cabling, and these have been categorized to represent the cable’s ability to transfer data quickly and over certain distances.
Most average people and small businesses don’t need to be transferring data at full gigabit speeds and would opt for what’s known as 100BASE-TX, or fast ethernet. This standard would be Category 5 (or better) and would support speeds of 100 Megabits per second over a length of about 100 meters.
An ethernet cable plugged into a router:
Not every device on a network would really need to be connected to the router via an ethernet cable. For example, a small business would probably connect its Point of Sale system (POS) via ethernet, as it is an essential workstation on the network. In a home, you would probably connect a personal computer that serves as a workstation via ethernet. We do this, because cables provide a more reliable and faster connection than simply using the wireless access point feature.
How does ethernet actually work?
The Media Access Control (MAC) Address
A device’s MAC address is assigned to it when manufactured. It is the physical address of a network adapter and is unique to each device. When reading a MAC address, it is displayed in hexadecimal format. This address, unlike an IP address is 48 bits, or 6 bytes, long and would look something like:
6D:C6:99:67:7F:02
The first half of this address is unique to the manufacturer and is known as an Organizationally Unique Identifier (OUI). The second half is unique to the device itself, and is actually its serial number.
The Ethernet Frame
In layer 2 of the OSI Model, the data link layer, ethernet handles frames. It would be taken care of by the switching functions of a SOHO router. These frames are started with a preamble, binary data used for synchronization. This preamble is then followed by the Start Frame Delimiter (SFD), which indicates the end of the preamble.
Following the SFD, we can find the destination MAC address of the device that this frame must be forwarded to as well as the source MAC address of where the traffic is coming from. After the destination and source, a frame would contain the EtherType, which describes the data in the next portion, the payload. The payload is the actual data being sent. Payloads contain layer 3 and higher data.
The frame is finally completed with a Frame Check Sequence (FCS). Essentially, this part confirms that the data was not corrupted while being sent. In the off chance that it was, this frame would not be forwarded to the device.
Logical design of a frame:
The Full-Duplex Ethernet
The topology, or physical design, of the network to create a SOHO LAN, would make a full-duplex ethernet. This means that traffic may be sent and received at the same time, without worrying about network collisions.
Previously, an ethernet topology may have included a hub, which would repeat all traffic to all devices connected to it. This is known as a half-duplex ethernet. The primary issue with this design is that if two devices were to communicate at the same time, a collision would occur, significantly slowing network speeds and efficiency. It also posed serious security issues as it would leave all devices vulnerable.
With the full-duplex, specific traffic may reach specific devices. This improved efficiency by only having the necessary devices communicate, which inherently made the network more secure by having only one device exposed to potentially malicious traffic, instead of every single one on the network. The full-duplex feature is thanks to the switching capabilities of modern routers.
To avoid these “collisions”, the full-duplex design deploys a protocol known as Carrier Sense Multiple Access/ Collision Avoidance (CSMA/CA). In general, this protocol would start by checking if the network is clear to send traffic. If it is clear, it would transmit a Ready To Send (RTS) signal; otherwise, it would prompt the device to wait for a random back off time until the channel clears. If the RTS signal is detected, the protocol checks for a Clear To Send (CTS) signal then goes ahead and transmits the frame to the appropriate device. If an RTS signal is detected, but not the CTS, the device would also be prompted to wait for a random back off time until both conditions are met.
Logical design of CSMA/CA:
Thank you for reading! Let me know if I missed something or if you learned anything new at: http://www.linkedin.com/in/colton-gabertan-463836209
