Student Name: Student Number: Faculty of Science Final Examination Computer Science 308-435B Basics of Computer Networks Examiner: Prof. Hans Vangheluwe Friday, April 27 th, 2001 Associate Examiner: Prof. Tim Merrett 9:00 12:00 INSTRUCTIONS: 1. Answer all questions directly on the examination paper. 2. No notes, books, calculators, computers or other aids of any type are permitted. 3. Translation dictionaries may be used. 4. The exam has 20 questions on 15 pages. The total number of page, including this cover page, is 16. 5. Attempt all questions: partial marks are given for incomplete but correct answers. 6. Numbers between brackets [] denote the weight of each question. The exam is out of a total of 105 points. 7. Use the back of the last page as scrap (it will be ignored during grading). The rear of the other pages may be used as extra space to answer questions. Good luck!
308-435B Final 1 (1) [20] What is the purpose of layering in networks? Give an overview of all the layers of the OSI model. For each layer, describe briefly its function. Give the relationship between the TCP/IP and the OSI network models. (2) [2] What is the bitrate for a signal in which 10 bits last 20 µs? What is the duration of a bit for a signal with a bitrate of 5 Mbps? 308-435B Final 1
308-435B Final 2 (3) [14] Assume a data stream of ten 0s. Encode this stream using the following encoding schemes. Draw the encoding as voltages in function of time. All encodings should use the same time scale. Assume the last voltage before the data stream started was positive. 1. Unipolar. 2. Polar NRZ-L. 3. Polar NRZ-I. 4. Manchester. 5. Differential Manchester. 6. AMI. 7. B8ZS. 308-435B Final 2
308-435B Final 3 Briefly describe relative advantages and disadvantages of the different methods. (4) [3] Give 3 types of signal impairment. How does the electrical specification for sending data in EIA-232 deal with each? 1. 2. 3. (5) [2] Using the Nyquist theorem, calculate the sampling rate for an analog signal with frequencies from 2000 to 6000Hz. What is the bandwidth of the original signal? (6) [3] Calculate the baud rate for a 6000bps bit rate transmitted using 4-PSK. Explain 4-PSK. 308-435B Final 3
308-435B Final 4 (6) [6] Given 3 data streams: A (1.5 Mbps), B (4/3 Mbps), and C (2/3 Mbps). Use synchronous Time Division Multiplexing with frames containing 6 1-byte slots + 1 framing bit per frame to transmit these data streams. 1. Which technique needs to be used to properly put the data in the slots? 2. What is the bitrate (in Mbps) required of the communication channel between MUX and DEMUX? Give a description of your derivation. (7) [3] Given m bits of data and a checksum of r bits, derive a necessary relation between m and r for the checksum to be able to correct 1-bit errors. How many redundancy bits are required to encode 1001101 using Hamming code? Derive the Hamming code for the above bitstring. Number and store bits from the right (lower order 308-435B Final 4
308-435B Final 5 bits) (same as Forouzan). (8) [4] What is a sufficient condition on a CRC generator polynomial G x for detecting an odd number of transmission errors? Prove this. Describe the CRC process (on sender and receiver side). (9) [4] In the absence of errors, draw the evolution of sender and receiver sliding windows. Both are of size 6. 308-435B Final 5
308-435B Final 6 How many different numbers are used for the sliding windows? Draw the evolution if the following sequence of messages is sent (DATA from sender to receiver, ACK from receiver to sender): DATA0, DATA1, ACK2, DATA2, ACK3, DATA3, DATA4, DATA5, ACK6. (10) [4] Given the Petri Net model in Figure 1 Describe the scenario which gives rise to duplicate ACKnowledgements. Draw the evolution over time of the communication between sender and receiver. Relate this drawing to a sequence of transitions 308-435B Final 6
308-435B Final 7 Figure 1: Petri Net model of stop-and-wait protocol firing in the Petri Net. On the Petri Net, add places/transitions to count the total number of lost frames. 308-435B Final 7
308-435B Final 8 (11) [4] In the presence of errors, two types of sliding window data link protocols are used. Which are they? Explain both briefly and point out differences. (12) [4] What is data transparency? 308-435B Final 8
308-435B Final 9 How is data transparency achieved in character oriented protocols (BSC)? How is data transparency achieved in bit oriented protocols (HDLC) (give an algorithm)? (13) [4] In HDLC, what is the purpose of I-frames, S-frames, and U-frames? 308-435B Final 9
308-435B Final 10 (14) [4] Explain how CSMA/CD handles medium access on a multipoint link. Derive a formula for the maximum (collision domain) length L of a CSMA/CD (e.g., Ethernet) network segment given minimum frame size F, signal propagation velocity V, data rate D, (15) [4] What is the difference between an Ethernet Switch and an Ethernet Hub? 308-435B Final 10
308-435B Final 11 (16) [4] Explain how TDM can be used for switching. (17) [4] Explain the similarities and differences between a Circuit-Switched Connection and a Virtual-Circuit Connection. 308-435B Final 11
308-435B Final 12 (18) [4] Explain the differences between repeaters, bridges, routers and gateways. For each of these, specify at (upto) which level of the OSI model they operate. 308-435B Final 12
308-435B Final 13 (19) [10] Explain Distance Vector Routing. Construct a small example (no more than 4 networks) to illustrate the algorithm. (20) [4] Explain the operation of the TCP protocol by means of the state transition diagram given in Figure 2. 308-435B Final 13
308-435B Final 14 Figure 2: TCP/IP State Transition Diagram 308-435B Final 14
308-435B Final 15 (20) [4] What is I/O multiplexing used for? Describe the different uses of the select() function. Which data structure is used to select sockets? 308-435B Final 15