ECE544 Mid-Term Page ECE 544 Computer Networks II Mid-Term Exam March 29, 2002 Profs. & M. Ott Instructions: This is a 2 hr, OPEN BOOK exam. (Only the textbook, Peterson & Davie, Computer Networks, A Systems Approach and 2 pages of notes are permitted.) Answer Q (40 points) and ANY THREE from Q 2-6 (20 points for each question). If you answer more than 3, your grade will be determined as Q + the best 3 from Q2-6. In addition to the correct final answer, credit will be given for logical development of each problem using network architecture diagrams, protocol definitions, timing diagrams, flow charts, pseudo-code, etc. where appropriate. Please be sure to state your assumptions, give concise explanations (no essays!), organize your work logically, and write with legible handwriting. Question : Network Architecture: A company with offices in three locations wishes to establish an IP-based broadband services data network with the following requirements:. Geographic locations: Atlanta, GA; Princeton, NJ and Palo Alto, CA. 2. ~00 employees at each location, each with a networked PC; 25% of these users have an additional laptop computer with wireless access. 3. A file server and compute server at each site. 4. Primary applications for PC users: email, www, ftp, mbone, NetMeeting (an IP-based multiparty teleconference application). 5. Estimated traffic per active user is 00 Kbps average and ~ Mbps peak. Design an IP network that provides both local-area and wide-area connectivity for this company. Your answer should concentrate on Layer 2 & 3 functionality and must provide a complete view of components and protocols used. Use the following outline:. A high-level drawing of the whole network, identifying each major networking component (including Ethernet or wireless LAN, ARP servers, internal routers/ switches, DHCP servers, Internet backbone, etc.). [0 pts] 2. Specify the protocol stack(s) used at each network node in the system for (a) control, including addressing, routing, QoS management, etc.; and (b) for data transfer from application-to-application (for each application listed above) 3. Explain how end-users addresses (MAC address, IP address, etc.) are assigned, by what entities and using what protocols. 4. Each of the 3 locations can be considered an IP intra-domain. How is intradomain routing done, by what entities and using what protocols? 5. What protocol is used for inter-domain routing (i.e. from one site to another) over the public Internet? Identify the network elements which use interdomain routing, explain how it works, and draw the routing protocol stack at each such node.
ECE544 Mid-Term Page 2 6. Would you recommend any special features such as IP QoS or multicast for this network? If so, indicate what network elements are needed to support these features and show what changes are required in the protocol stacks drawn in item 2 above. 7. Estimate the link bandwidth required for each local site s connection to the Internet, based on the ~00 users with traffic profiles specified above. [Make suitable assumptions for local vs. external traffic flows.] Question 2. Shared Media LAN:. Explain the basic principle of Ethernet local area networks. Why is the length of an Ethernet segment limited to ~500m? What is the reason for exponential backoff after a collision occurs? Sketch the typical throughput vs. offered traffic curve for Ethernet. 2. Explain the basic principle of 802.b WLAN. Why is collision avoidance used instead of collision detection? How does the RTS/CTS procedure help avoid hidden terminal problems? 3. Consider the Ethernet LAN with three users as shown in the figure below. Shared Media (Ethernet) U Prop del = U2 Prop del = U3 T/2 T/2 In the above, assume that users U, U2 and U3 transmit packets of fixed length 0T, where T is the end-to-end propagation delay. Assume for simplicity that packet arrivals and transmissions at terminals occur at virtual slot boundaries t=0,t,2t,3t,4t, nt. Also, assume that the shortest Ethernet packet (after collision) is T sec, and that backoff algorithms delay by kt units of time, where k=,2,3... Now suppose U, U2 and U3 wish to transmit a new packet at times t=0, t2=0 and T3=5T respectively. Draw the sequence of transmissions and retransmissions on the channel from t=0 to the time when all three packets P, P2, P3 have been transmitted successfully (assuming no other traffic on the LAN). Provide explanations for each channel event. Use the following tabular format to give your results, and explain each entry with numbered footnotes. [0 pts] Time T activity or state T2 activity or state T3 activity or state 0 Packet arrives and is Packet arrives and is Idle transmitted transmitted T 2T 3T
ECE544 Mid-Term Page 3 Question 3. Network: Consider the network shown below. The terminal TA connects to a server SA to download a video-on-demand movie. The application program at the TA first retrieves a program guide via TCP/AAL5 over a UBR VC, and then after the TA selects a movie, the SA sends an MPEG2 movie via UDP/AAL5 using a VBR VC.. Show the sequence of signaling events at each node in the network necessary to set up the path for (a) the program guide query, and (b) video delivery. Present your results in the form of a protocol timing diagram at each node, using the format shown in the figure. 2. Fill in the corresponding VPI/VCI forwarding table entries at each adapter and switch in the end-to-end connection. Use port numbers as indicated in the figure and assign VCI numbers arbitrarily in the range 28-256; set VPI=0. 3. Show how the AAL segments the first TCP query message into the cell format used for data transfer. [5pts] 4. Explain how quality-of-service (QoS) is set up for the video connection. Describe applicable traffic shaping, call admission control and scheduling functions. Port 2 Port 4 5 3 TA Port 2 4 SA Adaptor (TA) Port 3 Switch A 3 Switch B 2 Switch C Adaptor (SA) Format of VCI/VPI forwarding table at each switch: Input Port # VPI/VCI Output Port # VPI/VCI 2 0.37 4 0.2....
ECE544 Mid-Term Page 4 Question 4. Intra-Domain IP Routing:. Explain the basic principle of distance-vector (DV) routing using the network example given below. Describe the RIP packet format and explain how it is used to implement a DV routing protocol. [4 pts] R 2 4 5 R2 R3 R4 R5 2. For the network given above, give global distance vector tables (as in Table 4.5, Pg. 285 of Peterson and Davie, 2 nd ed) for each of the following: a. Each node knows only the distance to its immediate neighbors b. Each node has reported the information it had in its preceding step to its immediate neighbors c. Step (b) is repeated. [6 pts] 3. Explain the principle of link-state routing using the network example in Fig 4.. Describe the OSPF routing protocol and explain how it is used to implement a link-state routing protocol. [4 pts] 4. Give the steps in the forward search routing algorithm (as in Table 4.9, Pg. 297, Peterson & Davie, 2 nd ed) from initialization to completion for the network given above. [6 pts] Question 5. Quality-of-Service:. What are the four mechanisms used to provide quality-of-service in IP networks? Explain each how each of these mechanisms helps to guarantee QoS. Label the network given below to show where each of these QoS mechanisms applies. Discuss how end-toend QoS is ensured referring to the example network. [8 pts] CPE-Access Network Interface Access-Core Network i/f Access-Core Network i/f CPE-Access Network Interface Sender Router Router 2 Router 3 Router 4 Receiver Core Internet
ECE544 Mid-Term Page 5 2. Suppose a router has three input flows and one output. It receives the packets listed in the table below all at about the same time in the order listed during a period in which the output port is busy but all queues are otherwise empty. Give the order in which packets are transmitted, assuming: a. fair queuing [3pts] b. weighted fair queuing with flow #2 having weight 2, and flows #,3 with weight. [3pts] Packet Size Flow 00 2 00 3 00 4 00 5 90 2 6 200 2 7 0 3 8 50 3 3. Suppose a router has accepted flows with TSpecs as shown below, described in terms of bucket filters with token rate r packets per sec and bucket depth B packets. All flows are in the same direction, and the router can forward one packet every 0. sec. a. What is the maximum delay a packet can face? [3 pts] b. What is the maximum number of packets from the third flow that the router would send over 2.0 sec, assuming the flow sent packets at its maximum rate uniformly? [3 pts] TSpec table: Question 6. TCP: Flow # r B 37 0 206 2 4 42 4. Explain the role of a reliable end-to-end transport layer protocol like TCP. List and explain each of the services provided by TCP. Illustrate with traffic traces or protocol timing diagrams where appropriate. [0 pts] 2. Suppose a TCP connection has a window size of 8 segments, an RTT of 800 ms, a periodic sending rate of one segment every 00 ms, and the receiver sends back ACKs at the same rate with no delay. A segment is lost, and the loss is detected by the fast retransmit algorithm on receipt of the third duplicate ACK. At the point when the ACK of the retransmitted segment finally arrives, how much time has the sender lost compared to lossless transmission if: (a) the sender waits for the ACK from the retransmitted lost packet before sliding the window forward again (b) the sender uses the continued arrival of each duplicate ACK as an indication it may slide the window forward one segment. [0 pts]