Mid-Sem Examples

Question 1

Name the 7 OSI Layers, from Lowest to Highest

Answer 1

Application,
Presentation,
Session,
Transport,
Network,
Data Link
Physical

Question 2

List 2 advantages and 2 disadvantages of strictly conforming to a layered networking standard, such as OSI/ISO model

Answer 2

• (A) Layer corresponds to a different level of abstraction
• (A) Each layer corresponds to a well defined, independent function
• (A) Assures that users networking software will connect to other vendor’s system and software
• (D) Doesn’t allow creativity of other techniques
• (D) Exploitable for DoS hacks

Question 3

Which two layers would you employ encryption in for the OSI model?

Answer 3

Employ encryption at the lowest two levels because they’re the easiest to cause a DoS for.

Question 4

A system uses n-layer protocol hierarchy. Appl generates messages of M bytes each. At each layer, a header of h bytes is added. What fraction of the network B/W is filled with headers?

Answer 4

(n.h)/(m+n.h)

Question 5

Provide 3 distinct reasons why a stream of data is broken into frames by the DL layer?

Answer 5

• Easier to send small chunks of data in frames
• Can control the f;pw pf frames depending on traffic
• Detecting and correcting higher-level transmission errors from the network layer

Question 6

Reason for piggy-backed ACKs in the DL layer?

Answer 6

Used to decrease the number of frames sent back from the receiver to sender, by combining data and ACK in to one frame

Question 7

Why does the checksum only cover headers?

Answer 7

Without the header, the data is useless as it has no destination or source so would never be received by dest.

Question 8

Why is ACK preferred to NACk in stop-and-wait protocol?

Answer 8

• Timeouts would have to be longer to account for busy routes/lots of traffic
• ACK preferred as shorter timeout and sender knows sooner whether to resend data
• Less frames wasted if data is retransmitted quicker

Question 9

For the sliding window protocol, is it meaningful for the receiver’s max window size to be larger than the sender’s max window size?

Answer 9

Receiver’s window size should be constant so that the receiver doesn’t pick up unwanted frames as well as sender knowing when it sold receive ACK from receiver, knowing that the frames are in order

Question 10

Primary difference between Eth hub and Eth switch?

Answer 10

Hub receives packet and broadcasts to all outgoing ports but switch is more intelligent and only retransmits to correct/most efficient outgoing port.

Question 11

List 3 tasks a network bridge must perform connecting dissimilar hardware technologies

Answer 11

• Buffering of frames to allow for varying bit rates

- Data link reformatting for the 3 different standards of 802.3, 802.4 and 802.5

Question 12

Why must IEEE 802.3 employ collision avoidance instead of collision detection?

Answer 12

Wireless NIC cannot transmit and receive at the same time, therefore cannot detect a collision if transmitting. Employs a scheme to avoid collisions, MACA (multiple access with collision avoidance) using both physical and virtual channel sensing.

Question 13

Why are all MAC addresses in 802.11 NICs unique?

Answer 13

Unique identifier for a particular device

Question 14

Why is registering only wanted MAC addresses for an AP ineffective against a determined intruder?

Answer 14

Hackers are able to create fake MAC addresses so would be able to passively listen to packets to find out which MAC address is registered.

Question 15

Provide 2 distinct motivations for the NL to employ sequencing numbers

Answer 15

• Make sure all packets are received at dest from src for routing purposes so that the routing arg can be improved because nodes know which packet has been seen before
• In load shedding, better to discard higher seq num frames to descrease the num of frames kept in buffer.

Question 16

Why doesn’t each node in a virtual circuit use the same number to identify the circuit?

Answer 16

Each nodes needs to be able to alter outgoing ID numbers of packets from different computers so that they remain unique. VC doesn’t know which numbers are in use, the numbers only identify the SWITCHING between different nodes, and cannot identify whole circuit.

Question 17

Differences between flow control and congestion control

Answer 17

CC - ensures that the subnet can carry the offered traffic, which is a global issue for hosts and routers
FC - concerned with end-to-end control, possibly multiple hops apart

Question 18

Effects of Leaky Bucket and Token Bucket flow control

Answer 18

LB - ensures uncontrolled flow of packets into bucket, acts as a buffer for a regulated flow of bursty traffic out
TB - provides a bucket with tokens, with regulated flow where each packet must ‘consume’ a token before being transmitted.