Network topology refers to the physical layout (structure) of a network including cables, computers and other resources. This determines how components communicate on the network (LAN, WAN, MAN, CAN etc), affects network's performance and growth potential and determines type of equipment to purchase and how to manage network.
Local Area Networks (LANs) can be organized into various different structures. The most important thing is that there are physical and logical topologies. The physical topology is how wires are organized or run in the network (refers to arrangement of cabling) whereas the logical topology is how the network behaves logically, that is, how data or network messages travel between computers/nodes on the network.
Three basic LAN topologies exist, namely:
- Star (Hub-and-Spoke)
Sometimes referred to as linear-bus topology based on its structure, it is a simple design that utilizes a single length of wire/cable known as the "backbone", "bus" or "the medium" with both ends terminated with a terminator to absorb electronic signal traveling on the network. All computers/nodes on the network are connected to the backbone, by a transceiver either directly or by using a "drop cable" (normally a coaxial cable) and a BNC barrel connector.
With bus topology, only one computer can send information at a time. The information or data is broken into packets and sent as electronic signals that travel on the backbone, only the computer or node to which the data is addressed accepts it although all the other nodes on the network sees the transmission as signals move from point of transmission to both ends of the bus.
- Simple to set up.
- Failure of a device does not affect communication on the network
- Additional devices can be added anywhere on the bus
- Requires less cable than star topology
- Good for small networks and for quick or temporary LAN installations
- Inexpensive to implement on a small scale
- easy to add another workstation
- The length of the bus is limited by cable or signal loss
- This topology is very difficult to troubleshoot. Locating a break in the cable ("backbone"), or the device causing the fault, when the entire network is down can be time-consuming
- Management costs often too high
- Subject to congestion from network traffic
- The must be terminated at both ends to prevent signal bounce
Star (Hub-and-Spoke) topology
A star physical topology, used on a LAN, usually does not look like a star, except on paper. The focal point of this technology is on the centralized device (hub or switch) to which all the nodes or devices are connected. These nodes or devices are easily connected or disconnected using network media, such as UTP (Unshielded Twisted Pair) cable to the centralized device.
Each station only talks to the centralized device. During data transmission or communication, data is transmitted from the point of transmission (node sending data) to the centralized device. Normally, if a hub is used as the centralized device, the hub retransmits the data to all the connected nodes, devices or stations on the network but in the case where a switch is used as the centralized device, it is able to only send data to the destination station based on the packet address.
- Cabling is inexpensive and easy
- Network growth is easily accommodated
- Locating and repairing bad cables is straightforward
- Reliable and easy to administer and maintain
- If a centralized device (hub or switch) fails, the entire network on the centralized device fails.
- All nodes on the network receive the same signal, dividing the bandwidth
- The maximum UTP network cable length is 100 meters (about 330 feet).
- The maximum number of computers is 1,024 on a LAN
In a ring physical topology, network devices are wired and connected in a conceptual circle. A ring topology is almost always implemented in a logical ring topology on a physical star topology. Each device is attached to two other devices and uses the same network transmission signal, forming path in the shape of a ring.
Network data flow is unidirectional, and a controlling device, such as a hub or switch, intercepts and manages the data flow to and from the ring. Each device has a NIC (Network Interface Card) that contains a network transceiver, which both sends and receives signals. This topology uses network token-passing access methods referred to as Token Ring. Token Ring is the most common types.
- Only the device that holds the token can transmit packets of data, which eliminates network packet collisions
- Signal degeneration is low
- Difficult to troubleshoot and locate the problem cable in a network segment
- Hardware is proprietary and expensive
- Requires more network cables as compared to the bus topology
Topology and Medium Recommendation
COPPER WIRE CABLES
The copper wire cables is the oldest installed cables and are still the most used material for connecting devices. Almost every topology has its copper version. NCC Education Limited (2007, pg.15) describes in its Enterprise Networking coursework that:
'Copper wires are thinner with cable sheaths case. The original cables used paper for insulation and lead for the cable sheath, but modern cables use plastic insulation and polythene cable sheathing'. Copper wire cable have the below types:
- Twisted pair (Unshielded Twisted Pair, Shielded Twisted Pair)
- Coaxial cable
- Twinaxial cable
Advantages of copper wire cables
Cost effective/ Affordable: It is easy to acquire and at a reasonably affordable price as compared to the other transmission media. This affordability is the major reason why most developing countries of the world use copper wire for telecommunications transmission.
Easy installation: It is very easy to install or setup.
Disadvantages of copper wire cables
RADIO (WIRELESS) TRANSMISSION
According to Kimaldi (n.d), benefits of the wireless transmission media includes the below:
- Completes the access technology portfolio - Wireless enables a fully comprehensive access technology portfolio to work with existing dial, cable, and DSL technologies.
- Goes where cable and fiber cannot - the inherent nature of wireless is that it doesn't require wires or lines to accommodate the data/voice/video pipeline. As such, the system will carry information across geographical areas that are prohibitive in terms of distance, cost, access, or time. It also sidesteps the numerous issues of ILEC colocation.
- Provides broadband access extension - Wireless technologies as a transmission medium play a key role in extending the reach of cable, fiber, and DSL markets, and it does so quickly and reliably providing access in geographies that don't qualify for loop access and also commonly provides a competitive alternative to broadband wireline.
- Involves reduced time to revenue - companies can generate revenue in less time through the deployment of wireless solutions than with comparable access technologies because a wireless system can be assembled and brought online in as little as two to three hours.
Though the wireless transmission medium is replete with the advantages described above, it has its share of downfalls. Some identified disadvantages according to wirelesslans (2004) are:
Security - Many wireless networks utilize encryption technologies, but not all of these are effective. Comparing wireless with a wired network (copper wire cables), wireless technology offers much less security, since a hacker would have to tap into the actual cables in the case of a wired network.
Reliability - Like any radio frequency transmission, wireless networking signals are subject to a wide variety of interference, as well as complex propagation effects (such as multipath) that are beyond the control of the network administrator.
Speed - The speed on most wireless networks (typically 1-108 Mbps) is reasonably slow compared to the slowest common wired networks (100Mbit/s up to several Gbit/s).
Energy - Power consumption is fairly high compared to some other standards, making battery life and heat a concern.
Range - While sufficient for a typical home, it will be insufficient in a larger structure. Range varies with frequency band, as Wi-Fi is no exception to the physics of radio wave propagation.
FIBRE OPTIC CABLES
Fiber optics is the branch of science and engineering concerned with optical fibers. The optical fiber can be used as a medium for telecommunication and networking because it is flexible and can be bundled as cables. These fiber optic wires consist of advancements to the older copper wires and thereby replacing copper wires as the primary communication carriers in most developed countries. The wire's small size provides an advantage over copper wires. Fiber optic cables weigh less than copper wires, which makes installation easier and allows the wires to be placed in smaller areas (Palais, 1998). According to (hubtechinsider, June 4 2009), advantages of fibre optics include:
- Resistance to electromagnetic interference - Fiber has a very low rate of bit error (10 EXP-13), as a result of being so resistant to electromagnetic interference.
- Extremely high bandwidth - No other cable-based data transmission medium offers the bandwidth that fiber does.
- Early detection of cable damage and secure transmissions - Fiber provides an extremely secure transmission medium. By constantly monitoring an optical network and by carefully measuring the time it takes light to reflect down the fiber, splices in the cable can be easily detected.
- Easy to accomodate increasing bandwidth - Using many of the recent generations of fiber optic cabling, new equipment can be added to the inert fiber cable that can provide vastly expanded capacity over the originally laid fiber.
Some addition advantages of fibre optic according to (Bradley, n.d) includes:
- Durability - Fiber optic cable is less susceptible to breakage than other types of cable
- Distance - Fiber optic cable can carry a strong signal over a wide distance, resulting in a better quality transmission of voice and/or images.
Although fiber optics is renowned for their efficiencies and loads of advantages, there are a few drawbacks in them. Hubtechinsider(June 4, 2009a) states the below disadvantages:
- Susceptibility to physical damage - Fiber is a small and compact cable, and it is highly susceptible to becoming cut or damaged during installation or construction activities.
- Installation costs, while dropping, are still high - Despite the fact that fiber installation costs are dropping by as much as 60% a year, installing fiber optic cabling is still relatively costly as compared to the other media types.
- Special test equipment is often required - Equipment such as an OTDR (Optical Time Domain Reflectometer)is required, and expensive, specialized optical test equipment such as optical probes are needed at most fiber endpoints and connection nexuses in order to properly provide testing of optical fiber.
- System configuration - Converting existing hardware and software for the use of fiber optics does take a lot of time and money which also reduces the turnover for any profit making firm in the market
Limitations in Local Area Networks
In Local Area Networks, fiber optics is not used as widely as one would expect. One reason is the implementation requires great deal of changes in current networks and systems. This requires a lot of time and effort which the management is not willing to sacrifice. People are comfortable with what they have and don't want to change. Although most problems regarding program changing can be solved, the solutions to it will take much longer than expected. Thus, any new program has to be a big improvement over the old one to justify a significant change (although the great improvement usually means that the old program does not work).
Another fundamental problem in fiber optic LANs is the change in technology. The hardware and software to make LAN run efficiently add up to an expensive package. If many terminals in a building must be in constant touch with each other and a variety of other hardware, such as printers and storage devices, LAN will be cost efficient. However, if the real need is to keep the terminals in touch with a mainframe computer, it would be cheaper to run cables between them and the mainframe. If the terminals need to talk to each other, ordinary telephone lines could very well be used as telephone lines are much cheaper than fiber optics.
The major practical problem with fiber optics is that it usually costs more than ordinary wires. All costs elements involved in economic evaluation can be grouped into two main classes; which are investment costs and operation costs. The investment costs usually includes expenditures related to acquiring and owning properties and plants, in this case changing wires to fiber optic cables. All investment costs should be considered, such as those incurred for equipment and materials (also including storage and handling costs), engineering costs and miscellaneous costs. Operation costs include the usage of fiber optics and the wear and tear of it. The higher costs of fiber is often not by itself. Fiber optic cables are much cheaper than coaxial cables. The main difference comes when all the other components of fiber optics add up, such as transmitters, receivers, couplers and connectors. Fiber systems require separate transmitters and receivers because they cannot directly use the electrical output of computer devices; that signal must be converted into optical form and then converted back into electrical form. Fiber optic connectors and couplers are more expensive than any other electrical components. These costs are the ones that add up and form the major disadvantage of fiber optics.