Question $1$:

a$)$ Consider a network in which packets are delivered reliable across each link.

That each, for each pair of adjacent nodes A $-$ B$,$ the datalink layer has an ARQ

protocol $($such as stop$-$and$-$wait$)$ that ensures packets sent by A do arrive at B$.$

In this network, why do you think is necessary for the transport layer to also

have its own ARQ protocol and do retransmissions of its packets?

b$)$ Consider a network in which the data$-$link layer protocol between two

machines in the network was replaced by a different data$-$link layer protocol.

In general, how will this affect the transport, session, presentation, and

application layers?

Question $2$

$==========$

Assume we have two hosts A $-->$ B separated by a link. The propagation delay of

the link is $100$ms$,$ the bandwidth of the link is $2,000$ Kbytes$/$sec$,$ and all

packets are $1$ Kbyte. Assume the processing delay at B is $5$ ms$.$

a$)$ What is the bandwidth$-$delay product of the link?

b$)$ What is the transmission delay of a packet?

c$)$ What is the propagation delay of a packet?

d$)$ Assume that at time t$,$ host A begins to transmit the first bit of the

packet. At what time is the packet delivered to the application at B$?$

Question $3$

$==========$

Assume that the following $*$data$*$ bits are sent across a link

$1011000011010011$

what are the $*$physical$*$ bits $($low$,$ high,$)$ sent if:

a$)$ we use manchester encoding

b$)$ if we use $4$B$/5$B $($assume initially low$)$

c$)$ we use NRZI in addition to $($b$)$ above $($assume initially low$)$

Question $4$

$==========$

Consider the bit insertion protocol. There are several ways to encode "idleness"

a$)0111111111111111111111111110($i$.$e$.,$ a $0$ followed by more than $6$ ones$)$

b$)0111111011111111111111111101111110($i$.$e$.$ the $01111110$ flag, then many ones, then the flag again$)$

c$)01111110011111100111111001111110($i$.$e$.$ many flags in a row$)$

Argue that the receiver pseudocode $($that we gave in the slides$)$ will throw

away all of these bits $($i$.$e$.$ it will consider them idleness and throw them

away$)$

Question $5$

$=========$

Consider the parity bit protocol with column parity bits $($the P$\text{'}$s$)$ and row

parity bits $($the Q$\text{'}$s$).$ No $"$r$"$ bit in the corner. We said this protocol can catch

all single and double bit errors. PROVE it by doing a case analysis of where the

corrupted bits can be located.

Question $6-$ CRC and parity

$===========================$

a$)$ Assume that we are sending a message using CRC$,$ and an error occurs that

only corrupts data bits $($no CRC bits are corrupted, i$.$e$.,$ not bits in the

remainder R are corrupted$).$ Will the receiver catch all errors of this type?

If so$,$ explain why. If not, give a counter$-$example.

b$)$ Assume that we are sending a message using CRC$,$ and an error occurs that

only corrupts the CRC check bits $($i$.$e$.,$ only the bits in the remainder R$).$

Will the receiver catch all errors of this type? If so$,$ explain why. If not,

give a counter$-$example.

c$)$ Same as a$),$ except we are using parity, with the $"$p$\text{'}$s$"$ and the $"$q$\text{'}$s$"($no r$)$

d$)$ Same as b$),$ except we are using parity, with the $"$p$\text{'}$s$"$ and the $"$q$\text{'}$s$"($no r$)$

Question $7-$ CRCs

$=================$

Assume that our generator polynomial G is the product P$*($X$+1),$ where P is a

primitive polynomial $($i$.$e$.$ it does not divide $11,101,1001,10001,$ etc$).$

a$)$ Assume P$,$ as a polynomial, ends in $"+1".$ What are the types of errors that

G can detect? $($x$-$bit detection, where x is $1,2,$ or $3?)$

b$)$ Assume P$,$ as a polynomial, does not end in $"+1".$ What are the types of

errors that G can detect? $($x$-$bit detection, where x is $1,2,$ or $3?)$

Question $8$

$==========$

Consider the stop$-$and$-$wait $($i$.$e$.,$ alternating bit$)$ protocol. Modify the

protocol so that instead of obtaining the data $($i$.$e$.$ the body$)$ of the frame

from the "higher layer", we get it from an array of integers, and the receiver

has an array of integers where it stores the data received. The purpose is to

copy the array in the sender to the array in the receiver $($one element at the

time of course$)$

I.e$.,$ add the following array to the sender:

datasender : array $[$integer$]$ of integer;

ns: integer $/*$ initially $0*/$

I.e$.,$ an infinite array $($hence$,$ the index can be any integer$).$ ns is an index of

which data is the next one to send $($i$.$e$.$ next data to send is datasender$[$ns$]).$

In the receiver, add the following array

datareceiver : array $[$integer$]$ of integer;

nr: integer $/*$ initially $0*/$

nr is an index of where to store the next data to be received

frames will now have the format

frame$($b$,$ x$)$

where b is a bit $(0$ or $1)$ and x is an integer $($from the data array of course$)$

Question $9$

$==========$

a$)$ Consider the concurrent logical channels protocol. Could you add something

similar to part $($a$)$ above to the concurrent logical channels $($without major

modifications to the protocol$)?$ Explain why yes or why no$.$

b$)$ Consider again the concurrent logical channels protocol. Can we eliminate the

bit in the acknowledgment? I.e$.,$ can ack's b