清華大學 計算機網路 期末考

(+5)【 單選題 】
(5%) 1. Which of the following statements are correct for delivering packets (datagrams) over the Internet ?
(a) All packets destined to the same destination will be forwarded along the same routing path.
(b) The routing path of a packet is determined first before the packet is sending into the Internet.
(c) The Internet is reliable so that all packets will be received by the destination correctly.
(d) A packet may be partitioned into several fragments by routers.
(e) All fragments of a packet will be forwarded along the same path to the destination.
(+5)【 複選題 】
(5%)2. Which of the following statements are correct for “statistical multiplexing” ?
(a) Data are transmitted based on demand of each flow.
(b) Data are transmitted based on the arrival time of each flow.
(c) The bandwidth used by different flows may be different.
(d)The statistical multiplexing is fair due to the bandwidth used by different flows are the same.
(e)It provides better link utilization than TDM (Time-Division-Multiplexing) and FDM (Frequency-Division- Multiplexing).
(+5)【 複選題 】
(5%) 3. Which of the following are correct for CSMA/CD (Carrier Sense Multiple Access with Collision Detection) protocol ?
(a)Listen (carrier sense) before transmit.
(b)If channel is sensed idle, then transmit entire frame without considering if frame collision is detected.
(c)If channel is sensed idle, then transmit a frame and stop the transmission when a collision is detected.
(d)If channel is sensed busy, then defer transmission until the channel is idle.
(e)If channel is sensed busy, then defer transmission for a random time.
(+5)【 複選題 】
(5%) 4. The “collision window” of the CSMA/CD protocol is time required to detect a collision. Which of the following are correct for collision window ? Assume the one way propagation time of the network is “a”
(a)The collision window is equal to a
(b)The collision window is equal to 2a
(c)The worst case scenario happens when the two hosts are at opposite ends of the Ethernet
(d)The best case scenario happens when the two hosts are close to each other.
(e)The worst case scenario happens when the two hosts are close to each other.
(+5)【 複選題 】
(5%) 5. What are the purposes of RTS and CTS control frames used in IEEE 802.11 wireless LAN ?
(a)Reserve the wireless channel for a time period
(b)Informs all nearby nodes that a transmission is about to begin
(c)These two frames also contain data to transmit.
(d)The sender is asking the receiver to send back a frame after a time period
(e)The duration field in the RTS and CTS frames is used to specify the time period needs to transmit the frame
(+5)【 複選題 】
(5%) 6. Which of the following statements about the CSMA/CA are correct ?
(a)A station with a pending MPDU may transmit immediately when it detects the medium is free.
(b)A station with a pending MPDU may transmit when it detects a free medium for greater than or equal to a DIFS time.
(c)If the medium is busy when a station desires to transmit an MPDU, the station can transmit the MPDU when it detects a free medium for greater than or equal to a DIFS time.
(d)In CSMA/CA protocol, no packet collision will happen as the “collision avoidance” mechanism is used.
(e)If the medium is busy when a station desires to transmit an MPDU, the Random Backoff Time algorithm shall be followed after it detects a free medium for greater than or equal to a DIFS time.
(+5)【 複選題 】
(5%) 7. Consider the procedure to maintain the spanning tree of IEEE 802.1D spanning tree algorithm. For a bridge B with three ports x,y,z and root bridge ID = w, rpc = p. Assume port x is an R port, port y is a D port, port z is blocked and connected to a LAN M. Also assume port z has a transmission cost of c. What actions will be taken by the bridge B if the hello BPDUs suppose to be periodically received from port z are lost for a long time (a timeout event) ?
(a) Port z is selected as a new R port.
(b) Port z is selected as a D port.
(c) A BPDU (with root bridge ID = w, rpc =p) is forwarded to port z
(d) A BPDU (with root bridge ID = w, rpc = p+c) is forwarded to port z.
(e) A BPDU (with root bridge ID = w, rpc =p) is forwarded to port y.
(+5)【 複選題 】
(5%) 8. Consider the procedure to construct the spanning tree. When a bridge B (with root bridge ID = w, rpc = p) receives a BPDU (with root bridge ID = m, rpc =q) from port x (with a transmission cost of c). Assume w > m. Then what actions will be taken by bridge B ?
(a) port x will be selected as a new D port.
(b) port x will be selected as a new R port.
(c) A BPDU (with root bridge ID = m, rpc =q) will be forwarded to all the ports, except port x.
(d) A BPDU (with root bridge ID = w, rpc =q+c) will be forwarded to all the ports, except port x.
(e) A BPDU (with root bridge ID = m, rpc =q+c) will be forwarded to all the ports, except port x.
(+4)【 複選題 】
(4%) 9. Which of the following tagging rules for a “hybrid link” are correct ?
(a)For each VLAN, all frames traversing a particular hybrid link can be tagged in different ways.
(b)For each VLAN, all frames traversing a particular hybrid link must be tagged the same way: all implicitly tagged or all carrying the same explicit tag.
(c)There can be a mix of implicitly and explicit tagged frames but they must be for different VLANs.
(d)All the frames must be tagged explicitly.
(e)A frame traversing a particular hybrid will be tagged or untagged depends on which VLAN the frame came from.
(+6)【 填充題 】
(6%) 10. Consider the following VLAN configuration. For switch 1, what are the “member set” and “untagged set” of each VLAN ? please answer the port “number” for each of the following questions.
注意:在每一個VLAN中的答案請依數字大小排序,答案請由"," 隔開,並且答案欄裡不要輸入任何空白字元(答案後面也不要有空白)
EX:若A空格答案有4 2 1 3,請填入:1,2,3,4(注意逗號是英文輸入法的逗號,若不確定請直接複製 ",")
Member set:
VLAN1:__A__
VLAN2:__B__
VLAN3:__C__
 
Untagged set:
 
VLAN1:__D__
VLAN2:__E__
VLAN3:__F__
A=ANS:2,3,4,5
B=ANS:7,11,12
C=ANS:6,7,8,9
D=ANS:2,3,4,5
E=ANS:11,12
F=ANS:8,9
(+5)【 複選題 】
(5%)11. Let SeqNumToAck denote the largest sequence number not yet acknowledged, such that all frames with sequence number less than SeqNumToAck have been received. Which of the following statements are correct for the “cumulative acknowledgement” ?  
(a)The receiver acknowledges SeqNumToAck+k to sender if a frame with sequence number of SeqNumToAck + k, k > 0, is received.
(b)The receiver acknowledges SeqNumToAck to sender if a frame with sequence number of SeqNumToAck + k, k > 0, is received.
(c)The receiver then sets LFR = SeqNumToAck + 1 and LAF = LFR + RWS-1.
(d)The receiver then sets LFR = SeqNumToAck and LAF = LFR + RWS.
(e)The receiver then sets LFR = SeqNumToAck - 1 and LAF = LFR + RWS.
(+5)【 單選題 】
(5%)12. Let MaxSeqNum be the total number of available sequence numbers.  Assume RWS = SWS, then which of the following relationship is sufficient to distinguish between different frames of the same sequence number ? 
(a)SWS + 1 ≤ MaxSeqNum
(b)SWS ≤ (MaxSeqNum+1)/2
(c)SWS = (MaxSeqNum+1)/2
(d)SWS < MaxSeqNum x 2
(e)SWS < (MaxSeqNum+1)/2
(+6)【 填充題 】
(6%)13. With the CIDR, what is the “prefix” for all the networks 192.4.16 through 192.4.31 ? Assume the prefix is represented as         192.4.a/b, and therefore
      a = _______A_______
      b = _______B_______
 
A=ANS:16
B=ANS:20
(+4)【 複選題 】
(4%)14. Which of the following statements are correct for “Link State” routing protocol ? Assume that each node knows the cost of the link to each of its directly connected neighbors. 
(a)Each node constructs a one dimensional array (a vector) containing the “distances” (costs) to all other nodes and distributes that vector to its immediate neighbors.
(b)Each node constructs a one dimensional array (a vector) containing the “distances” (costs) to all other nodes and distributes that vector to all other nodes.
(c)Each node constructs a one dimensional array (a vector) containing the “distances” (costs) to its immediate neighbors and distributes that vector to its immediate neighbors.
(d)Each node constructs a one dimensional array (a vector) containing the “distances” (costs) to its immediate neighbors and distributes that vector to all other nodes.
(e)The link state packets (LSP) are sent to all other nodes by using reliable flooding algorithm.
(+5)【 複選題 】
(5%)15. Which of the following statements are correct for “flow control and congestion control” ?
(a)Flow control involves preventing senders from overrunning the capacity of the receivers
(b)Flow control involves preventing senders from overrunning the capacity of the networks.
(c)Congestion control involves preventing senders from overrunning the capacity of the networks.
(d)Congestion control involves preventing senders from overrunning the capacity of the receivers.
(e)Congestion control involves preventing senders from overrunning the capacity of the receivers and networks.
(+5)【 單選題 】
(5%) 16. For the Additive Increase Multiplicative Decrease (AIMD) mechanism used in TCP congestion control, TCP does not wait for an entire window’s worth of ACKs to add 1 packet’s worth to the congestion window, but instead increments CongestionWindow by a little (denoted as increment) for each ACK. Which of the following is correct for this increment ?
(a) Increment =1/CongestionWindow
(b) Increment = 1/MSS
(c) Increment = MSS/CongestionWindow
(d) Increment = MSS (MSS/CongestionWindow)
(e) Increment = MSS/(CongestionWindow+MSS);
(+5)【 複選題 】
(5%) 17. Which of the following statements are correct for “slow start” in TCP congestion control ?
(a) Slow start effectively increases the congestion window linearly
(b) Slow start is used to increase the congestion window rapidly from a cold start
(c) When the ACK for a packet arrives TCP source, the TCP source adds 1 packet size of MSS to CongestionWindow
(d) When the ACK for a packet arrives TCP source, the TCP source doubles the size of CongestionWindow
(e) Slow start effectively doubles the number of packets the TCP source has in transit every RTT
(+5)【 複選題 】
(5%) 18. Which of the following statements are correct for a TCP source to handle the “packet lost” in congestion control ?
(a) After receives 3 duplicate ACKs, the CongestionWindow is cut in half and then grows linearly.
(b) After receives 3 duplicate ACKs, the CongestionWindow is set to 1 MSS and then grows linearly.
(c) After timeout event happes, the CongestionWindow is set to 1 MSS; the window then grows exponentially to a threshold, and then grows linearly.
(d) After timeout event happes, the CongestionWindow is set to 1 MSS; the window then grows linearly.
(e) After timeout event happes, the CongestionWindow is cut in half; the window then grows exponentially to a threshold, and then grows linearly.
(+10)【 填充題 】
(10%) 19. Consider the following figure to demonstrate the growth of CongestionWindow size for different “packet lost” events in TCP congestion control. We can use the following events to “describe” the curve in this figure. Initially, the connection is in slow start period with CongestionWindow size = 1 (1MSS).
        Event 1: 3-duplicate ACKs happens at 5th RTT
        Event 2: Timeout happens at 10th RTT
        Event 3: 3-duplicate ACKs happens again at 18th RTT.
        And we have a = 16 (window size), b = 8 (window size), c = 12 (window size), d = 6 (Threshold), e = 10 (window size), f = 5       (Threshold) as shown in the figure.
 
  Now, consider we have the following events:
        Event 1: 3-duplicate ACKs happens at 4th RTT
        Event 2: Timeout happens at 17th RTT
        Event 3: 3-duplicate ACKs happens again at 25th RTT.
        Then we have 
a = _____   (window size), 
b = _____   (window size), 
c = _____   (window size), 
d = _____   (Threshold), 
e = _____   (window size), 
f = ______  (Threshold).
 
And what is the CongestionWindow size at the following points?
At 2nd RTT :__________g______
At 18th RTT :__________h_____
At 23th RTT :__________i______
At 29th RTT :__________j______

a=ANS:8
b=ANS:4
c=ANS:16
d=ANS:8
e=ANS:12
f=ANS:6
g=ANS:2
h=ANS:1
i=ANS:10
j=ANS:9