Network Topology and The Internet
A string of bits that are sent as signals, and a set of connections scattered across the borders — there lies the idea of the Internet.
What is a Network Topology? Network topology is the arrangement of the elements (links, nodes, etc.) of a communication network. Topologies are either physical (the physical layout of devices on a network) or logical (the way that the signals act on the network media, or the way that the data passes through the network from one device to the next)[1]. Network topology can be used to define or describe the arrangement of various types of telecommunication networks, including command and control radio networks, industrial field busses, and computer networks.
Note: To illustrate a concrete example to physical and logical topologies, you can think of them as being wired (there is a physical layout) or being wireless (the signal act on the network media).
While network topologies can be applied into arrays of knowledge domain such as social networks, ontology models, and genomics, our focus herein is only limited (while the core concepts still hold and can be applied into other fields) to telecommunication systems which, in turn, hold significance to our understanding with the Internet — it is also important to note that the concepts we are building in this article are more closely related to the construction of the internet (with small “i”), nonetheless, understanding these concepts will build our foundations to comprehend the complexities of the Internet.
For all intents and purposes, I will leave a few notes to build our intuition with some notable concepts that have lead to the Internet.
1.1. Fundamental Elements of a Graphs: A brief overview of graph theory and network topologies
The abstraction of network topology is graph theory, that is, network topologies applies the notions established in graph theory.
What is graph theory: In discrete mathematics, graph theory is the study of graphs, which are mathematical structures used to model pairwise relations between objects. A graph in this context — does not pertain to a graph of a function (typically projected into some coordinate space) rather is made up of vertices (also called nodes or points) which are connected by edges (also called links or lines).
Whereas graph theory has a ton of stuff to offer into molding our basic knowledge of these topologies, our focus [in this article] is to build our understanding with some of the basic structures (topologies) of networks. So, this article will not discuss graph theory, in detail — if ever you are interested and want to build a deeper understanding with these networks I would recommend reading this article https://bit.ly/2yjWM5M.
Suffice to say, that the notion of vertices, edges, undirected (is a graph in which the two endpoints of each edge are not distinguished from each other.) and directed graphs (is one in which the edges have a distinguished direction, from one vertex to another) will help us understand the core concepts behind these topologies — introduced in the succeeding sections.
The very basic building block of the World Wide Web is built upon the concepts introduced here.
1.1.1. Bus topology (linear topology)
Bus topology is a network type in which every computer and network device is connected to a single cable. It transmits the data from one end to another in a single direction.
Note: No bidirectional feature is in a bus topology, meaning the data transmits only in one direction.
The Big Idea:
A host[2] on a bus network is called a station. In a bus network, every station will receive all network traffic, and the traffic generated by each station has equal transmission priority. A bus network forms a single network segment[3] and a collision domain[4]. In order for nodes to share the bus, they use a media access control technology such as carrier sense multiple access (CSMA) or a bus master.
Advantages
● It is easy to connect a computer or device and typically it requires less cable than a star topology.
○ Very easy to connect a computer or peripheral to a linear bus.
○ Requires less cable length than a star topology resulting in lower costs
○ The linear architecture is very simple and reliable
○ It works well for small networks
○ It is easy to extend by joining cable with connector or repeater
○ If one node fails, it will not affect the whole network
Disadvantages
● The entire network shuts down if there is a break in the main wire and it can be difficult to identify the problem if the network shuts down.
○ The entire network shuts down if there is a break in the main cable or one of the T-connectors break
○ A large number of packet collisions on the network, which results in high amounts of packet loss
○ This topology is slow with many nodes in the network
○ It is difficult to isolate any faults on the Network
1.1.2. Ring topology
Ring topology forms a ring connecting devices with its exactly two neighboring devices. Furthermore, a ring network (ring topology) is a network topology in which each node connects to exactly two other nodes, forming a single continuous pathway for signals through each node — a ring. Data travels from node to node, with each node along the way handling every packet.
Note: Rings can be unidirectional, with all traffic traveling either clockwise or anticlockwise around the ring, or bidirectional (as in SONET/SDH).
Advantages
● It can span larger distances than other types of networks, such as bus networks because each node regenerates messages as they pass through it.
○ A very orderly network where every device has access to the token and the opportunity to transmit
○ Performs better than a bus topology under heavy network load
○ Does not require a central node to manage the connectivity between the computers
○ Due to the point to point line configuration of devices with a device on either side (each device is connected to its immediate neighbor), it is quite easy to install and reconfigure since adding or removing a device requires moving just two connections.
○ Point to point line configuration makes it easy to identify and isolate faults.
○ Reconfiguration for line faults of bidirectional rings can be very fast, as switching happens at a high level, and thus the traffic does not require individual rerouting.
Disadvantages
● Aggregate network bandwidth is bottlenecked by the weakest link between two nodes.
○ One malfunctioning workstation can create problems for the entire network. This can be solved by using a dual ring or a switch that closes off the break.
○ Moving, adding and changing the devices can affect the network
○ Communication delay is directly proportional to the number of nodes in the network
○ Bandwidth is shared on all links between devices
○ More difficult to configure than a Star: node adjunction = Ring shutdown and reconfiguration
1.1.3. Star topology
A star network is an implementation of a spoke–hub distribution paradigm[5] in computer networks. In a star network, every host is connected to a central hub. In its simplest form, one central hub acts as a conduit to transmit messages[6]. The hub can be passive in nature i.e., not intelligent hubs such as broadcasting devices, at the same time the hub can be intelligent known as active hubs. Active hubs have repeaters in them.
Advantages
● One malfunctioning node does not affect the rest of the network.
○ If one node or its connection breaks, it does not affect the other computers nor their connections
○ Devices can be added or removed without disturbing the network
○ Works well under heavy load
○ Appropriate for a large network
Disadvantage
● If the central computer fails, the entire network becomes unusable.
○ Expensive due to the number and length of cables needed to wire each host to the central hub
○ The central hub is a single point of failure for the network
1.1.4. Mesh topology
A mesh network (or simply meshnet) is a local network topology in which the infrastructure nodes (i.e. bridges, switches, and other infrastructure devices) connect directly, dynamically and non-hierarchically to as many other nodes as possible and cooperate with one another to efficiently route data from/to clients. This lack of dependency on one node allows for every node to participate in the relay of information. Mesh networks dynamically self-organize and self-configure, which can reduce installation overhead. The ability to self-configure enables the dynamic distribution of workloads, particularly in the event a few nodes should fail. This, in turn, contributes to fault-tolerance and reduced maintenance costs.
There are two types of mesh topology: Full mesh topology — which occurs when every node has a circuit connection it to every other node in a network — and partial mesh topology — which is less expensive to implement, and also yields less redundancy.
Note: Mesh topology may be contrasted with conventional star/tree local network topologies in which the bridges/switches are directly linked to only a small subset of other bridges/switches, and the links between these infrastructure neighbors are hierarchical. While star-and-tree topologies are very well established, highly standardized and vendor-neutral, vendors of mesh network devices have not yet all agreed on common standards, and interoperability between devices from different vendors is not yet assured
1.1.5. Hybrid topology
A hybrid topology is a collection of two or more topologies that are described above. This is a scalable topology that can be expanded easily. It is a reliable one but at the same time, it is a costly topology.
1.2. Protocols
For a network to function reliably, it is important to establish rules which control their behavior. These rules are called protocols.
Note: there are lots of different protocols — developed and adopted protocol standards to build products for network applications — designed to be compatible with other products of vendors.
In this article, we shall limit our discussion to the most common and important protocols for bus topology and for star topologies namely CSMA/CD, and the CSMA/CA.
1.2.1. Carrier Sense, Multiple Access with Collision Detection (CSMA/CD) Protocol
This protocol dictates that each message be broadcast to all the machines on the bus (see Figure 4.2). Each machine monitors all the messages but keeps only those addressed to itself. To transmit a message, a machine waits until the bus is silent, and at this time it begins transmitting while continuing to monitor the bus.
Note: CSMA/CD is not compatible with wireless star networks in which all machines communicate through a central AP. This is because a machine may be unable to detect that its transmissions are colliding with those of another.
1.2.2. Carrier Sense, Multiple Access with Collision Avoidance (CSMA/CA) Protocol
The most common approach to collision avoidance is based on giving an advantage to machines that have already been waiting for an opportunity to transmit. The protocol used is similar to Ethernet’s CSMA/CD. The basic difference is that when a machine first needs to transmit a message and finds the communication channel silent, it does not start transmitting immediately. Instead, it waits for a short period of time and then starts transmitting only if the channel has remained silent throughout that period.
1.3. Combining Networks
Sometimes it is necessary to connect existing networks to form an extended communication system. This can be done by connecting the networks to form a larger version of the same type of network.
1.3.1. Building a large bus network from smaller ones
In the case of bus networks based on the Ethernet protocols, it is often possible to connect the buses to form a single long bus. This is done by means of different devices known as bridges (or repeaters), and switches, the distinctions of which are subtle yet informative. Bridges connect a smaller network segment from a large network — it relies on LAN to LAN connection which is shown as the link between two buses. While switches connect more network segments than bridges.
1.3.2. Enter internet
Note: internet is different from The Internet which we shall discuss briefly in the succeeding section.
Sometimes, however, the networks to be connected have incompatible characteristics. For instance, the characteristics of a WiFi network are not readily compatible with an Ethernet network. To wit, the networks must be connected in a manner that builds a network of networks, known as an internet, in which the original networks maintain their individuality and continue to function as autonomous networks.
The connection between networks to form an internet is handled by devices known as routers, which are special-purpose computers used for forwarding messages. Note that the task of a router is different from that of repeaters, bridges, and switches in that routers provide links between networks while allowing each network to maintain its unique internal characteristics.
Final notes
Now that we have built our basic knowledge with network topologies, network protocols, and build our basic understanding with the internet, I will have to leave some few notes and key concepts that extend our understanding into building the intuition behind the Internet (the world wide web of networks).
The idea of the world wide web of networks can be briefly described in the figure below where Internet Service Providers (ISPs) are built on top of each other forming a hierarchy of networks. These networks are then organized by spatial regions and sub-regions forming a collection of clusters connected on top of each other — that’s the Internet.
Summary
We have learned that network topologies are an application of rules grounded by graph theory. Further, we have reviewed some of the basic network topologies and provided a summary of their advantages and disadvantages; we reviewed some of the basic protocols and build our understanding of the internet. Lastly, with the aid of ISP’s network architecture, we build our intuition to the workings inside the Internet (the world wide web of networks).
Remarks for this article
These contents are mainly derived from Brookshear, J. G. (2008). Computer science: an overview. Addison-Wesley Publishing Company.
References:
[1] Physical topology is the placement of the various components of a network (e.g., device location and cable installation), while logical topology illustrates how data flows within a network.
[2] A network host is a computer or other device connected to a computer network. A host may work as a server offering information resources, services, and applications to users or other hosts on the network.
[3] A network segment is a portion of a computer network.
[4] A collision domain is a network segment connected by a shared medium or through repeaters where simultaneous data transmissions collide with one another.
[5] The spoke-hub distribution paradigm is a form of transport topology optimization in which traffic planners organize routes as a series of “spokes” that connect outlying points to a central “hub”.
[6] Roberts, Lawrence G.; Wessler, Barry D. (1970), “Computer network development to achieve resource sharing”, AFIPS ’70 (Spring): Proceedings of the May 5–7, 1970, spring joint computer conference, New York, NY, USA: ACM, pp. 543–549, DOI:10.1145/1476936.1477020.
[7] Brookshear, J. G. (2008). Computer science: an overview. Addison-Wesley Publishing Company.
[8] Cui, Kainan & Xiaolong, Zheng & Zeng, Daniel Dajun & Zhang, Zhu & Luo, Chuan & He, Sk. (2013). An Empirical Study of Information Diffusion in Micro-blogging Systems during Emergency Events. 7901. 140–151. 10.1007/978–3–642–39527–7_16.
