Sessional 1
1 of 4
Sessional 1 Solution
Component 1: Section 1
True/False
1. False
2. False
3. False
4. True (if window size=1), False otherwise
5. True
6. True
7. False
8. False
9. False
10. True
Component 1:Section 2
Short Answers
1. Since in packet switched networks the resources are not reserved along the route
from source to destination so the upper bound on delay is not guaranteed. The
service can severely degrade in time of congestion.
2. [There are various camps which support different techniques to support video in
the statistically multiplexed networks e.g. one is that there must be classes of
network traffic given different priority (its like ensuring or allocating resources in
need basis).]
In TCP/IP stack, video is supported using UDP.
3. Packet-switch buffer is full
Packet is corrupted (most DLL protocols check the integrity of the frame so it can
drop at packet-switch/intermediate nodes)
Packet is duplicated (packet is dropped at receiver)
Receivers buffer is full
4. In the case of GBN, n packets/frames have same timeout timer (simple:
advantage), since it uses cumulative acknowledge in case of timeout sender has to
send all the previously unacknowledged packets/frames (disadvantage). In
selective repeat SR each packet has its own timeout timer (complex:
disadvantage). In the case of timeout only the unacknowledged packet is sent
(disadvantage). Generally if there is high rate of packet loss and bit errors, then
SR is better than GBN.
5. Layers implement protocols. There has to be a protocol on which a layer works so
they are not mutually exclusive. Also a “thing” cannot be a protocol and a layer at
the same time so they are mutually exclusive in this respect.
docsity.com
