YEAR 108

I. [50%] ( 2 25 )

1. In a complete circular DHT of 212 peers with identities from 0 to 4095, each node has shortcuts to its

next 2k -distance neighbors of (0k11) and will forward a message to its maximum-distance neighbor

as possible. To send a query from Peer 51 to Peer 1292, which is NOT the distance traveled along the

path? (A) 8 (B) 16 (C) 32 (D) 64

2. Which is functioned by a DASH server? (A) Divides a video file into multiple chunks (segments). (B)

Encodes each chunk at different rates. (C) Provides IP address resolution to a given domain name of

a URL. (D) Returns a manifest file that provides URLs for different versions of a video along with

different bit rates.

3. A RDT2.2 (NAK-free) sender is in the state of waiting for ACK 0. It will resend the data packet if it

receives (A) ACK 0 (B) NAK 0 (C) ACK 1 (D) timeout

4. A RDT3.0 (channels with loss) sender is in the state of waiting for ACK 0. It will resend the data

packet if it receives (A) ACK 0 (B) NAK 0 (C) ACK 1 (D) timeout

5. In a TCP server application, which socket call will return the connection socket for a new incoming

connection? (A) accept() (B) bind() (C) listen() (D) socket()

6. Which is NOT supported by UDP? (A) in-order (B) best effort (C) connectionless (D) multiple

processes

7. Which must have a buffer to receive out-of-order data? (A) go-back-N (B) selective repeat (C) TCP

(D) stop-and-wait

8. Which is NOT related to the set of TCP timeout interval? (A) estimated RTT (B) sampled RTT (C)

sampled timeout (D) sampled deviation

9. Which header field is included in UDP but not in TCP? (A) acknowledgement number (B) sequence

number (C) checksum (D) length

10. Which header field is included in IPv4 but not in IPv6? (A) time-to-live (B) protocol (C) checksum

(D) length

11. An acknowledgment for a packet with sequence number n indicating that all packets with a sequence

number up to and including n have been correctly received at the receiver is called (A) negative

acknowledgement (B) selective acknowledgement (C) cumulative acknowledgement (D) duplicate

acknowledgement

12. Which is the mechanism applied when the congestion window is below the Threshold and the window

grows exponentially? (A) AIMD (B) fast recovery (C) fast retransmit (D) slow start

13. For the same expected seq #, which does a TCP receiver send the highest acknowledgement #?

(A) Arrival of in-order segment with expected seq #. All data up to expected seq # already ACKed.

(B) Arrival of in-order segment with expected seq #. One other segment has ACK pending.

(C) Arrival of segment that partially or completely fills gap.

(D) Arrival of out-or-order segment higher-than-expected seq #. Gap detected.

14. Which is the state of a TCP server lifecycle when it receives SYN and sends ACK? (A) LAST_ACK

(B) TIMED_WAIT (C) SYNSENT (D) SYNRCVD

15. Which is the state of a TCP client lifecycle to let it resend the final acknowledgment in case the

ACK is lost? (A) LAST_ACK (B) TIMED_WAIT (C) CLOSE_WAIT (D) FIN_WAIT_2

16. Which host is NOT within the same network of subnet /22? (A) 192.168.3.1 (B) 192.168.4.1 (C)

192.68.5.254 (D) 192.168.6.254

17. What is the network mask for subnet /23? (A) 255.255.0.0 (B) 255.255.254.0 (C) 255.255.255.0 (D)

255.255.255.128

18. Which technology reduces the need of global unique IPv4 addresses? (A) ICMP (B) NAT (C) CIDR

(D) DHCP 19. Which technology increases the provision of global available sub-networks? (A) ICMP (B) NAT (C)

CIDR (D) DHCP

20. Which device always needs to examine the higher layer header of an incoming packet than the others?

(A) NAT box (B) bridge (C) router (D) hub

21. For file distribution, the P2P model is expected to perform better than the client-server model by

utilizing (A) server download bandwidth (B) server upload bandwidth (C) peer download bandwidth

(D) peer upload bandwidth

22. A browser establishes two parallel TCP connections to the same Web server. When the host running

the browser receives a segment from the server, which field of the segment is different? (A) source IP

address (B) destination IP address (C) source TCP port number (D) destination TCP port number

23. When a DNS server receives two queries from the same host, which field of the query message is

different? (A) source IP address (B) destination IP address (C) source UDP port number (D)

destination UDP port number

24. All UDP datagrams leaving a local network via a NAT box must have the same (A) source IP address

(B) destination IP address (C) source UDP port number (D) destination UDP port number

25. During TCP three-way handshake, which of the following segment does the server send initial

sequence number and buffer size with? (A) SYN (B) SYNACK (C) ACK (D) FIN

II. [50%] ( 50 )

1. (3%/3%/4%) Consider reliable transport-layer services.

(a). In our rdt protocols, why did we need to introduce timers? ($3, R10)

(b). In our rdt protocols, why did we need to introduce sequence numbers? ($3, R9)

(c).Comparing with GBN, what are the special features of the sequence numbers in TCP?

2. (3%/3%/4%) Consider network-layer services.

(a). What are the three typical types of switching fabrics?

(b). What is the role of the forwarding table within a router? ($4, R4)

(c). Repeating with (b), how does the router utilize the table for forwarding a packet?

3. (10%) Consider an idealized case of two hosts, one located on the West Coast of the United States and

the other located on the East Coast. The speed-of-light round-trip propagation delay between these

two end systems, RTT, is approximately 30 milliseconds. Suppose that they are connected by a channel

with a transmission rate, R, of 1 Gbps and the size of a packet is 1,500 bytes, including both header

fields and data. How big would the window size have to be for the channel utilization to be greater

than 98 percent? ($3, P15)

4. (10%) Consider the given network of six nodes, and assume that each node

initially knows the costs to each of its neighbors. Use Dijkstras algorithm

to compute the shortest path from node a to all the other network nodes.

Demonstrate how the algorithm works by computing a table step-by-step

and show the final forwarding table with paths and costs to the others.

5. (10%) Consider the given network of four nodes, and assume that each node

initially knows the costs to each of its neighbors. Use the distance vector algorithm

to compute the distance table of each node step-by-step. Besides, show the final

forwarding table with paths and costs to the others from node w.

d

c

a

e

f

6 12

8 5

1 3

2

b

7

4

w

y

z

x

1

3

2

4

6