Understanding Network Topology And Communication Protocols

Network Topology

The overall physical or logical design of computer networks is what is referred to as a network topology. In essence, they outline the pattern of devices connected to a network. 

Star Topology – A simple design where all computers are usually connected to a central server or device e.g. a hub or switch. Due to its simplicity, it’s the most commonly used pattern (Zandbergen, 2017). E.g. computers usage in an office set up. Nevertheless, during operations which inherently is usually communication, data always flows through the central device to the pre-determined destination. Therefore, the central device acts as a management unit that coordinates operations.

1. Fast to install because of its simplicity.
2. High fault tolerance
3. An optimum design that can easily detect faults.

1. It’s expensive to install.
2. Easily affected by failures if the central device is damaged.
3. Bus Topology – Characterised by a central network or connecting cable known as the bus. Examples are lab computers accessing the same physical cable. Now, during operations, data packets are sent across the bus reaching all the devices attached to it. However, only the intended recipients processes the data and only after confirmation it’s the intended destination. It’s therefore a good model for small networks e.g. a classroom or lab as resource utilization affects the overall performance as well as the issues involved (Certification kits, 2017).

1. Easy to install.
2. A simple pattern that requires few resources to set up.

1. A collaborative topology that cannot function on its own.
2. When the central cable is affected, it’s hard to identify the root of the problem (Omnisecu, 2017).
3. Mesh Topology – A connection pattern where all device is connected to each other forming a continuous redundancy. The best example being the internet. In general, two types of mesh designs exists; full and partial mesh. In partial structure, devices are not fully connected where some devices have independent peers as compared to others. On the other hand, full mesh is characterised by absolute connectivity, where all devices are peered with each other (Certification kits, 2017).

1. Robust and very resilient to faults.
2. Easy to add new devices with little effect on the overall network.
3. Its overall design can handle large volumes of data (Nizam, 2014).

1. Hard to maintain.
2. It tends to require more resources to set-up. 

In the telecommunication industry, several devices are used, these devices vary in design and structure which necessities the need for standards to govern their operations. The TCP/IP model is a hierarchical protocol that is made up of different layered modules that function in collaboration to meet the need of communication (Omnisecu, 2017).

 When transmitting or receiving data the devices involved must outline several variables including items such as the location and the size of data transmitted. Control information is a data bundle that is added to the actual data to regulate and direct it. Encapsulation will occur at each layer of the TCP/IP model where different control information is added at each stage forming an extended data package (Omnisecu, 2017). In reverse, during decapsulation, control data is used to extract information in the receiving device. 

From the diagram above, one can see the general procedure followed where, control data (header files) are added at each subsequent stage. Moreover, they become part of the overall data packet at each stage where an additional header file is added at the subsequent layer.

As compared to the processes discussed above, multiplexing and de-multiplexing involve the transmission of multiple signals or data streams across a single (or common) channel. Therefore, in multiplexing, several data streams will use a common channel (single) to move from one source to another. On the other hand, de-multiplexing is the process used to separate these data streams on the receiving end outlining each individual signal (Nizam, 2014).

So, we have: Bandwidth 6.8 MHz and SNR 132

From the bit rate formula (Nyquist): Bit rate = 2 x Bandwidth x Log₂ Signal level (L)

Moreover, Shannon capacity formula states: Capacity (Shannon capacity) = bit rate = B x Log₂ (1 + SNR). Therefore,

 B x Log₂ SNR = 6800000 x Log₂ 132

TCP/IP Model

Bit rate    = 47901880.01 = 47.9019 MHz

From the bit rate formula identified above (Nyquist) we have; Bit rate = 2 x Bandwidth x Log₂ Signal level (L)

Signal level is thus: 47901880.01 = 13600000 x Log₂ L

3.52219706 = Log₂ L, therefore, L = 2^3.52219706

= 11.48912529 Levels

When comparing the two, OSI (Open systems interconnection) model and TCP/IP model it’s good to outline their definitions and purposes. OSI model is a general communication standard that is used to define the interactions seen in communication systems (Burke, 2017). On its part, TCP/IP is a networking standard that is used to define the communication structure of devices across the internet. Therefore, while OSI model will have an extended use, TCP/IP will only apply when an internet connection is imminent. In fact, from its definition, IP will outline the processes used to acquire information in a given network while TCP identifies the channels used.

From this definition, it’s easy to see that the OSI model has extended functionalities as compared to the TCP/IP model. This outcome is a direct effect of the development timeline seen between the two where TCP/IP was created at a time when there were minimal system considerations. However, many of the communication protocols used today rely on the structure of the TCP/IP model to function. Consider protocols such as SMTP (Simple mail transfer protocol) they all are based on the TCP/IP model which outlines its importance and the reason to why OSI model has not taken over (Frenzel, 2013). 

1. A simple and generic model that outlines modules (layers) based on their functionalities, services, interfaces and protocols.
2. An adaptive model that supports all communication services.
3. In addition to this, it poses an abstract design that makes each layer robust and independent in functionalities.

1. It never outlines any operational protocols.
2. It’s also difficult to introduce new protocols as it’s an old model
3. All layers depend on each other to function.

1. It’s very easy to install because it operates independently of the system’s operations.
2. It’s a general protocol model that defines crucial communication protocols such as routeing protocols.
3. Furthermore, it’s a scalable model.

1. It has complex design thus is difficult to set up.
2. Moreover, it’s a slow model as compared to other recently defined models such as IPX.
3. In addition to this, it requires extended resources to function because of it overhead count (Jayasundara, 2017).

We are given: frame size 5 million bits, no. of routers10, the queuing and processing time 3.5 µs and 1.8 µs respectively, the length of the link 1900 km, c as 2.2 x 10⁸ m/s and the bandwidth 8 Mbps

Delay is therefore given by:

Processing + queuing + transmission + propagation time

Thus: Processing time = 10 x 1.8µ = 1.8 x 10^-5 s (accounting for 10 routers)

Queuing time = 10 x 3.5µ = 3.5 x 10^-5 s

Transmission time = Frame size/Bandwidth

= 5000000/8000000 = 5/8, approximately 0.625 s

Propagation time = Link length/speed of light

= 1900000/2.2 x 10⁸ = 8.636364 x 10^-3 s

Latency = 1.8 x 10^-5 + 3.5 x 10^-5 + 0.625 + 8.636364×10^-3 = 0.63369 s

Transmission time is therefore more dominant because it has a bigger packet size. On the other hand, Processing and queuing time are negligible (smallest values).

There are several protocols that are used to transmit and receive messages i.e. mails, POP3 (Post office protocol) is one such protocol that is precisely used to access remote servers hosting mailboxes. These remote servers will always retain the user messages and will only avail them when requested. Moreover, POP3 has limited functionalities unlike other protocols because it will only receive user messages (mail).

Operation: POP3 operates using four major states; authorization, transaction, update and closed (NCC, 1998).  

The states:

1. Authorization – The first state exhibited by the POP3 session where it must acquire the right authentication to access the servers. In this state, communication is initiated by the server which sends a greeting message to the client to signal the start of operations. As a response, the client will avail the verification information (authentication) so that it can be granted access to the mailbox stored in the server.
2. Transaction – When verified (authentication information), the client is able to access the mailbox as well as conduct various operations within it. For instance, the client can view and list all the messages stored in this location. Furthermore, its will also designate messages that are set for deletion.
3. Update – A consequent of the client operations which after completing its activities in the previous section will issue a QUIT command signalling the end of the operations. After the QUIT command is given, the POP3 session automatically enters the update state. Moreover, in this state, the server will also automatically delete all messages identified for deletion in the previous state by the server. The final operation is usually the termination of the TCP connection between the server and client.
4. Closed – An overall state that signifies the absence of operations. Closed state is usually exhibited before the authorization state when the client has zero access to the server.  

References

Burke. J. (2017). What is the difference between TCP/IP model and OSI model? Tech target. Retrieved 01 May, 2017, from: https://searchnetworking.techtarget.com/answer/What-is-the-difference-between-OSI-model-and-TCP-IP-other-than-the-number-of-layers

Certification Kits. (2017). CCNA – Bus, Ring, Star & Mesh Topologies. Retrieved 01 May, 2017, from: https://www.certificationkits.com/cisco-certification/ccna-articles/cisco-ccna-physical-networking-concepts-layer-1/ccna-bus-ring-star-a-mesh-topologies/

Chaudhari. A. (2016). 12 Advantages and Disadvantages of OSI model Layered Architecture. CSE stack. Retrieved 01 May, 2017, from: https://www.csestack.org/advantages-disadvantages-of-osi-model-layered-architecture/

Frenzel. L. (2013). What’s The Difference between the OSI Seven-Layer Network Model and TCP/IP? Electronic design. Retrieved 01 May, 2017, from: https://www.electronicdesign.com/what-s-difference-between/what-s-difference-between-osi-seven-layer-network-model-and-tcpip

Jayasundara. M. (2017). Advantages and disadvantages of TCP/IP and OSI model. Retrieved 01 May, 2017, from: https://msccomputernetworks.blogspot.co.ke/2016/08/advantages-and-disadvantages-of-tcpip.html

Kozierok. C. (2005). POP3 General Operation, Client/Server Communication and Session States. The TCP/IP guide. Retrieved 01 May, 2017, from: https://www.tcpipguide.com/free/t_POP3GeneralOperationClientServerCommunicationandSe-2.htm

Netscape Communications Corporation. (1998). Receiving Mail with POP3. Messaging Access SDK Guide. Retrieved 01 May, 2017, from: https://docs.oracle.com/cd/E19957-01/816-6027-10/asdk5.htm

Nizam. A. (2014). Advantages and Disadvantages of Using Mesh Topology. Networking basics. Retrieved 01 May, 2017, from: https://www.networking-basics.net/mesh-topology/

Omnisecu. (2017). TCP/IP Data Encapsulation and Decapsulation. Retrieved 01 May, 2017, from: https://www.omnisecu.com/tcpip/tcpip-encapsulation-decapsulation.php

Zandbergen. P. (2017). How Star, Bus, Ring & Mesh Topology Connect Computer Networks in Organizations. Retrieved 01 May, 2017, from: https://study.com/academy/lesson/how-star-topology-connects-computer-networks-in-organizations.html