EE450-SampleII-Nazarian-Summer11

EE450-SampleII-Nazarian-Summer11 - EE450 Sample Exam II...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: EE450 Sample Exam II Nazarian Student ID: _______________ Name: ____________________ Score: /90 Total Time: 65 minutes IMPORTANT: SHOW YOUR WORK. 1) TDM (Time Division Multiplexing) (27 points) Neda, a Physical layer designer at Motorola, designs a simple synchronous TDM Mux, which reads 5 bits of data from each input at a time. There are 250 input channels, however in average, only 10% of the inputs are active at a time. Each active input can generate data at a rate of 0.8Kbps. Neda considers 10 framing bits to be added to each output frame by the TDM process. a) (8 pts) Each input channel is originally analog with baseband frequency of 10Hz to 50Hz. Neda implements an A/D converter [Sampler+Quantizer+Encoder] for each input channel to transform the analog signal to digital before applying TDM on the channels. What is the sampling rate? and the rate at the of output the Quantizer? How many quantization levels the A/D must have? b) (4 pts) What is the length of each TDM frame at the output of the Mux? c) (6 pts) Calculate the output frame rate? d) (4 pts) What is the output rate in Mbps? e) (2 pts) Considering the overhead calculate the efficiency of this TDM. f) (3 pts) Neda needs to report to the customer the Mux performance which is defined as the useful data bits transmitted per second. Calculate the Mux performance and explain why it is greater than, less than or equal to the output rate you calculated in part (d). 2) Error Detection using CRC/FCS (13 points) Host A needs to send dataword 11010101 to host B. It uses the CRC division with divisor 11010 at its Data Link layer and sends the codeword as part of a frame. a) (1 pt) How many bits are in the code word? b) (8 pts) Show the CRC division process done at host. It’s your choice to use binary or polynomial representation of numbers. c) (1 pt) What is the codeword sent from A to B? d) (1 pt) Explain whether or not it is possible for the received codeword (at B) to be 111101000101? e) (1 pt) Explain whether it is possible that the CRC process at B checks a codeword with one or more bits of error and it detects no error? f) (1 pt) Explain whether or not B’s Data Link layer might pass the data part of a faulty codeword to its Network layer? 3) HTTP (10 points) The web browser is used in host A to obtain a webpage that is 4 MB (mega bytes) in addition to the basic HTML file, includes 7 objects. The IP address of the web server is cached already. The response time between the host A and the server is denoted by RTTA. Assuming zero transmission time of the object, how much time elapses from when the client clicks on the link until the client receives the object? a) (4 pts) How long does it take for the client to receive the webpage, using persistent HTTP without pipelining (i.e., no parallel TCP connection)? Justify your answer. b) (4 pts) Repeat part (a) for non-persistent HTTP with pipelining. c) (2 pts) Which case between (a) and (b) has a higher throughput? Calculate the throughput values for each part (a) and (b) to validate your answer. 4) Latency and Switching Technologies (30 points) In this problem we would like to derive a general formula for the total end-to-end delay for a TCP/IP network (which uses connectionless packet switching technology.) Let us consider sending M packets of equal length from host A to host B which are connected by N routers, R1, R2, ..., and RN. • All links have equal length and speed. The propagation delay, for each link is denoted by dprop (propagation delay from A to R1, from Ri to Ri+1, or from RN to B. • For each node the average processing+queuing delay to process a packet is denoted by dpq. This means a node can start transmitting the next packet as soon as it’s done transmitting the current one. All the nodes use the same data rate, R (bps). Each packet has P bytes. Therefore transmission delay of any node and for each packet can be denoted by dtrans =(8P)/R. a) (20 pts) Sketch the timing diagram and specify all delay components for M=4, N=3. It means 4 packets of equal size will be sent from A to B through 3 routers. • A R1 R2 R3 B Time b) (5 pts) Following part (a) write the total end-to-end delay from host A to send those 4 packets to host B. c) (5 pts) Let us define the performance (throughput) of the network as the amount of data (in bits) sent from A to B per unit of time. Write the performance formula based on the transmission of the 4 packets described in parts (a) and (b). 5) Stop&Wait ARQ Protocol (10 points) 10 frames are transmitted using the Stop&Wait Automatic Repeat Request protocol. The 5th, 7th and 8th frames are received with error and the 9th frame is dropped. They all are received with no errors detected when transmitted for the second time. Draw a timing diagram to illustrate the frame transmission between the sender and receiver nodes. ...
View Full Document

This note was uploaded on 08/27/2011 for the course EE 450 taught by Professor Zahid during the Summer '06 term at USC.

Ask a homework question - tutors are online