Candidates must answer THREE questions

Question $1$

$($a$)$ The local borough council needs to set up two computer networks using different structural

models. One network is to serve the recycling department which needs to have a good

resource sharing function to interact with various related services. The other network is to

serve for the council's documentation and archive sections which require high security and

good data back$-$up$.$

$($i$)$ What structural model would you advise to be employed by the borough council

for each department? Explain why.

$($ii$)$ Draw schematically the suggested network configurations.

$($iii$)$ What are the main advantages and disadvantages of peer$-$to$-$peer and client$-$server

networking?

$($b$)$ In a CSMA$/$CD network the transmitting station is attempting to retransmit data using the

truncated exponential backoff algorithm after a collision was detected.

$($i$)$ Explain what a truncated exponential backoff algorithm is and how it works.

$($ii$)$ If the minimum frame size of the network is $650$ bytes and the network is running

at $100M\frac{b}{s},$ what are the minimum and maximum waiting times for a station

before it will begin to retransmit again after $n=7$ successive collisions have

occurred?

$($c$)$ In computer networking the Data Link Layer is divided into two sub$-$layers $-$ the Logical

Link Control $($LLC$)$ and the Medium Access Control $($MAC$).$

$($i$)$ Explain the reasons for introducing the LLC and MAC sub$-$layers, and list the main

functions of these sub$-$layers.

$($ii$)$ Consider the LLC which uses the flag bits with bit stuffing method for data

framing. If the sending LLC receives the stream of data shown in Figure $1\mathrm{.}1,$ write

down the stream of the data as they appear on the line after the framing with the bit

stuffing.

$00011111011111011111111111111101111000$

Figure $1\mathrm{.}1$

Question $1($continued$)$

$($d$)$ An IPv$4$ network shown in Figure $1\mathrm{.}2$ consists of two hosts H$1$ and H$2$ and two routers R$1$

and R$2.$ The link H$1-$R$1$ has maximum transmission unit $($MTU$)$ MTU$1=1220$ bytes, the

link R$1-$R$2$ has MTU$2=610$ bytes, and the link R$2-$H$2$ has MTU$3=604$ bytes. The

network uses a non$-$transparent fragmentation. The application layer of the host H$1$

generates $1080$ bytes of data which are given to the transport layer for a reliable

transmission to the host $H2.$

$($i$)$ Is fragmentation of the packet required in this case? Explain your answer.

$($ii$)$ In case the packet is fragmented, calculate the payload of each fragment.