A network is a connection between one or more nodes (devices) regardless of distance and size. They are established for the sake of communication of data between the nodes. To accomplish this, the nodes will utilize circuit switching, message switching, or packet switching to send the data through from one signal to another. They are the basis for telephone networks, computer networks, and the Internet as a whole.

Types of Nodes

Nodes (or devices) are basically defined as any device on a network but they are actually a lot more than just that. There exist various different types of nodes and each type has a different purpose and function in a network which makes it important to identify them and know the difference between them.

A network interface controller (NIC) is a type of node which grants any computer the ability to access transmission media and process low-level network information. Basically, this hardware component is what allows a computer to access a network as it is what responds to traffic across a network which is intended for a specific network address or the NIC itself. Within an Ethernet network, each NIC will have a Media Access Control (MAC) Address which is usually stored in the device’s permanent memory. The Institute of Electronics and Electrical Engineers (IEEE) maintains and administers these MAC Addresses to ensure uniqueness across all of them. Each address is six octets (one octet = eight bits) big with the three most significant octets used to identify the NIC’s manufacturer. The manufacturer then assigns the three least significant octets using only their own uniquely assigned octet prefixes.

A repeater is a device with the purpose of receiving a signal from another node, cleaning any noise associated with that signal, and then retransmitting the signal at a higher power level. Noise and attenuation on a signal can cause the signal to be weaker and make it more difficult for nodes to connect to a network. This is especially likely if the signal has to travel through dense obstacles or over long distances. As such, repeaters are ideal for building networks which span over large distances even though they require a small delay in boosting the signal. This delay can affect the performance of the network and proper functionality of the network. An Ethernet hub is a repeater with multiple ports and although they were once common, they’ve been mainly replaced by switches in the modern world.

A network bridge connects and filters the transmissions between two networks so that they are joined together to form one. This would maintain a network’s unified broadcast domain whenever a network sends out a broadcast signal to all nodes on it. Conversely, some may choose to break down a network into smaller two or smaller segments which could make them less congested and therefore; more efficient.

A switch is a node which filters and forwards frames on between different ports on a network based on the MAC Address of the intended recipient of each frame. While an Ethernet hub can only broadcast the data it receives to all nodes connected to it, a switch is capable of recognizing and identifying which node connected to it is the one the frame is intended for thanks to the MAC Address attached. It will then learn to associate these MAC Addresses connected to these physical ports in the future and in the event it receives a frame with a MAC Address it is not familiar with, it will broadcast the message to all ports but the source to identify who it is intended for. Although the term may be used loosely in to refer to other nodes like router and bridge, a switch is indeed different as it can be considered a smart multi-ported bridge.

A router is a node which is capable of forwarding data between one network and another by processing the routing information in each packet of data sent. They have their own routing table which is used when forwarding data to another network and they are typically used for networks spanning huge distances. There are cases of a routing table having a null value which means that if the packets are sent to that routing number, they are dropped as nothing happens after they are sent.

A modem, so named from its task as a MOdulator-DEModulator, is used to connect nodes which were not originally designed for digital network traffic via wire or wireless hardware. This is accomplished by one or more carrier signals being modulated by the device’s digital signal and produce an analog signal which can then be altered to give the required attributes it needs for transmission. This process can be reversed when needed as well. The device is typically used by telecommunication companies using Digital Subscriber Line (DSL) technology. A firewall acts as a controller for network security and access on a network. Their purpose is usually to reject requests from unrecognized and unauthorized nodes while allowing requests from recognized and authorized nodes. A terminator is a type of node which may be placed at the end of a network to ensure that the signal sent there is killed off as oppose to be reflected which would create unwanted noise and interference on the network.

Types of Networks

The topology of a network is what ultimately determines the structure and type of network. It should be noted, however, that a network can also be identified not just by its topology, but by its geographical size. For the most part, there exist seven types of networks based on topologies and seven different types based on a network’s geographical size.

Starting first with the topologies, a point-to-point network topology is the simplest of the eight and is simply a single connection between two nodes on a network. A great example of this is a toy a child might have where they build a makeshift phone with two tin cans and a string adjoining them. The benefits of this topology is that because it is so simple, the signal between the two nodes cannot be impeded as it a direct connection, which also makes it the cheapest and easiest network to build.

The bus topology builds upon the point-to-point variation by adding in some complexity. It allows for more than two nodes to be connected to it by using the main cable which each node is connected to through interface connectors. This main cable can be referred to as the backbone of the topology and is called the bus; hence, the name. When a signal from a node on the network enters the bus, it travels both ways in a broadcast to all other nodes until it finds the intended recipient. Unfortunately, if the address of the recipient it finds does not match the intended address for the data, the data is ignored and considered a drop. While the bus topology might be cheap as it runs on just the one cable, it also has a high chance of failure because if the bus is damaged, the entire network fails.

The star topology goes a step further by having a central hub with point-to-point connections to every other node on the network. Following this logic, one could say that every node on the network is indirectly connected to each other thanks to the central hub. It should be noted that the hub here does not have to be an Ethernet hub as it can be also a switch or router as well. In this topology, the central hub functions as a server which acts as a signal repeater while every other node acts as a client as all data on the network must pass through the central hub. While this is a very scalable technology when it comes to adding more nodes to the network, it also has a very high chance for failure since if the central hub fails, then the entire network will also fail. It should also be noted that another flaw with the topology is that since all data must go through the central hub, it can become congested and thus become a bottleneck.

The ring topology is basically the bus topology going around in a circle (as it also known as the circle topology). Data will travel around the loop in one direction until it reaches its intended recipient with the nodes between the sender and the recipient repeating the signal each time so that it does not degrade. While the advantages of this topology is that there is no need for a network server to handle the connectivity between nodes and that as its size increases, so too does its performance, it is still possible for a bottleneck to occur on the weakest link between any two nodes. Also, if any of the links should break, the entire functionality of the network will fail.

A mesh network builds upon the idea of a point-to-point network by ensuring every single node on the network is connected to every single other node. While this is great for ensuring direct communication between every node and even ensuring that there’s always some means of communication regardless of if a link is broken, it is extremely costly with every node on the network.

A daisy-chain is a network that connects two computers together; or rather, from one computer to the next. Through this method, the signal will constantly pass from one system to the other until it finds its intended recipient. It can be designed either in a linear form, where there is a two-way link between each system (save for the ones at the ends of the chains) or in a ring form, where the two end-computers are joined together by another link and the topology adopts a logic similar to the ring topology. While daisy-chaining is the easiest way to expand a network (unless it’s a star topology) and offered some security in the event a link broke, it could become expensive the longer the chain grew.

Finally, there is the hybrid network which is simply any combination of the aforementioned topologies in such a way that it does not resemble any one of them. This type of topology is usually the most common as it allows flexibility when creating the networks to gather the pros of multiple topologies all at once.

Alongside the different topologies, there are also the different types of networks categorized by the geographical size which the network spans across. The smallest of these networks is called a personal area network (PAN) which is used for a computer communicating to other nodes which are within close vicinity to a person; such as a printer. The network can include wired and wireless devices and typically spans no more than 10 meters.

A local area network (LAN) is bigger than a pan and are about the size of a home, school, office building, or close cluster of buildings. Each device on the network is considered a node and the wired versions are done through Ethernet technology. It can be used as a residential area for a home area network (HAN).

A metropolitan area network (MAN) is a network which covers the size of about a university or college and can also be used for a college for a campus area network (CAN) as a result.

A wide area network (WAN) is much larger than a MAN, however; a LAN will offer higher data speeds (up to 100GBits/s). WANs are typically used to cover the scale of a city, country, or even intercontinental distances as it makes use of media such as telephone lines, cables, and airwaves.

An enterprise private network (EPN) is a network whose size is quite variable as it is determined by the size of the organization it is used. As such, the network can be as big as a LAN, or even a WAN that spans across countries. Typically, they are used so the company can share its resources between all of its employees regardless of their geographical location.

A global area network (GAN) is a massive network used to support an arbitrary number of LANS, WANS, satellite coverage areas, etc. This network is typically done for mobile carriers as the key issue for them is handing off communications from one coverage area to the next. A virtual private network (VPN) is one which is an overlay network that stands above another. As such, much like with an EPN, it’s size is determined by the size of the network underlying (for example, the Internet). They’re typically used by individuals to by-pass geo-blocked websites or data encryption when using a network.

Functionality and Applications

As the main goal of a network is for communication between nodes, there is a lot which has already been achieved with it and more which will be achieved with it in due time. The biggest achievement of all is the Internet which is simply the largest network in existence. Aside from that, telephones and cellular phones also utilize their own networks based on their mobile providers and companies utilize them to help streamline their workflow. As time progresses, there are even some exotic technologies being developed in hopes of going even further with networks, such as the Interplanetary Internet which extends the internet between planets through the use of radio waves.