L-3rr-1ICSE Date: 1410112017

BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY, DHAKA

10-3([-1 B. Sc. Engincering Examinations 2015-2016
Sub: CSE 307 (Software Engineering and Information System Design)
Full Marks: 280 Time: 3 Hours
The figures in the margin indicate full marks.
USE SEPARATE SCRIPTS FOR EACH SECTION

S~:CTION -A
There are FOUR questions in this section. Answer any THREE.

I. A library keeps records of current loans of books to borrowers. Each borrower is'
identificd by borrower numbcr and each copy of a book by an accession number. The
infonnation held about books is the title, author's name/s, publisher's name, publication
date, international standard book number (ISBN - a unique book identifier), purchase
price, classification (reference or fiction), and number of pages. A given book may be
written by a number of different authors. A book may cover a number of different
subjects. When a member of the library issues a borrow request, s/he is granted it if the
book is available and his/her personal borrowing restriction is not violated. Each member
has a restriction of maximum number of books to be allowed to borrow at a time
depending on the type of membership, e.g., student/teacher. When a book is borrowed,
the return date is automatically recorded based on the current date and the borrower's
elassifieation. Other borrowers, pending their return, may reserve books out on loan.
Borrowers who hold overdue books or who have reached their loan limit, arc flagged to

prevent further borrowings. (15+ 18+ 13 %)

(i) Design a Class diagram to implement Library system described above.
(ii) Design Collaboration diagram for the book borrowing scenario.
(iii) Draw a State Diagram for an Object of Book elass showing the associated
actions.

2. (a) Suppose we are designing an Accounting Software. We have a elass named Account
to represent the accounting heads in the chart of accounts. In each transaction one.
account is debited and at the same time another account is credited. There are elasses .
called Debtor and Creditor which stores the information which account is debited and
which account is credited. An account may have children. The account that does not have
any child is called leaf account. An account that has any transaction associated with it
cannot be deleted. Also an account which has child cannot be deleted. (18)
Draw the sequence diagram of an account
, deletion usc case where the user is first
presented with a list of accounts and user can select one to delete. If the deletion is
successful the user will be notified and if it is not possible then the user will be shown a
message saying why it is not possible.
(b) Discuss white-box testing in brief. What arc the advantages and disadvantages of
whi te-box testing?

(c) Discuss how co/!fidentially, integrity and non-repudiation arc implemented for
software applications. ' (16) .
Contd P/2.
Vi

.,
.. ' ;.' .
 =2=
CSE 307

3. (a) The requirement specification of a web-based software application stated that

maximum 100 users may access simultaneously. After deployment the application failed
to response properly when 90 users accessed at a time. It was recovered later and then on
a certain occasion 150 users tried to access and again it crashed. Which two software
testing methods could prcvent those two failures respectively? (8)
(b) Discuss MVC architecture with advantages and disadvantages. What indirect benefits

are offered by Code Review practice in a software development company?

(c) [fa student gets mark less than 50, s/he gets F grade. For 51-60 slhe gets C, for 61-70
gets B, and for marks obtained more than 70 he received grade A. Mark less than 0 is
invalid. Write a java method to implement JUnit testing that will take mark as input and
test the program that computes grade with the rule discussed above. Check for all
boundary values as well as invalid input. (16)

4. (a) A news agency gathers new and publishes them to different subscribers. You need to
create a framework for an agency to be able to inform immediately, when event occurs,
its subscribers about the event. The subscribers can receive the news in different ways:
Emails, SMS, etc. The solution need to be extensively enough to support new types of
subscribers (may be new communication technologies will appear). What will be the
appropriate design pattern to implement this scenario? Draw the diagram and write Java
code to implement the solution. (20)
(b) Briefly discuss eight common bad practices found in software development from the
security viewpoint. (12)
(c) What should you keep in mind to avoid merge conflict in version controlling systems?

What are the advantages of distributed version control system? (14 %)

SECTrON-B
There are FOUR questions in this scction. Answer any THREE.

5. (a) What are different layers in software architecture? Briefly explain. (10)
(b) In terms of use-case modeling, define the "extends" relationship with example. (10)
(c) Distinguish between thin client and fat clicnt. (6%)
(d) The Pizza Ordering System: (20)
The Pizza Ordering System allows the user of a web browser to order pizza for home
. delivery. To place an order, a shopper searches to find items to purchase, adds items one
at a time to a shopping cart, and possibly searches again for more items. When all items
have been chosen, the shopper provides a delivery address. If not paying with cash, the
shopper also provides credit card infonnation. The system has an option for shoppers to
register with the pizza shop. They can then save their name and address information, so
that they do not have to cntcr this information every time that they place an order.
Develop a use case diagram, for the Pizza Ordering System. The system should allow a
user to register and log in, and PaybyCredit, which requires credit card payments.

Contd P/3

'.
 •

=3=

CSE 307

6. (a) What is payback analysis? Explain briefly how it works.

(b) Distinguish between distributed and centralized architecture. What are the
disadvantages of distributed architecture? (12)
(c) Consider the following problem description: (20)
A major penal centre (PC) has been built outside the town of Dhaka for keeping prisoners
convicted of only one offense of white collar crimes. The prison facility has a constant
flow of prisoners into and out of the prison. They are also moving prisoners within the
prison based on their good behaviour. On a day to day basis, approximately 136 prisoner
changes take place. The changes are processed in the prison control centre office by
Control Centre Officer (CCO). Each day, the new prisoner processing division receives
the new prisoners, conducts a physical examination, assigns the prisoners to living
quarters and sends the information file on the new prisoners to CCO's office. CCO adds
infonnation on the new prisoner to a prisoner information database kept on the PC. CCO
also updates the prisoner locator log which keeps records of where each prisoner resides.
Finally, CCO files the actual folder away in an enormous storehouse of file cabinets
which contain information on all prisoners who have ever stayed at Dhaka prison. If a
new prisoner is found to have been a previous occupant of Dhaka prison, CCO
consolidates both files. As prisoners stay at Dhaka prison, the officials review their
behaviour record. Good behaviour or closeness to release time warrant an upgrade in
accommodations, usually to minimum security housing. Movement of prisoners to new
quarters is done on a weekly basis. Orders are issued to move the prisoners and the move
information is sent to CCO. CCO makes these changes in the prisoner locator log and the
prisoner information database. CCO also pulls the prisoners long term file and notes good
behaviour commendations. A release review and parole board reviews prisoner records
on a daily basis and generates a set of prisoners to be released either into the custody of a .
parole officer or without any restrictions. They notify the prisoner and send an update of
the release to CCO's office. CCO removes the prisoner from the prisoner information
database and prisoner locator log and update the long-term file of the prisoner to reflect
the release.
Draw a data flow diagram for describing the functional requirements ofthe above system.
(d) What is discovery prototyping? (4)

7. (a) What is software validation? Describe the software testing phases with necessary
diagram. (10)
(b) Discuss about the phase of classic life-cycle model. Write down the advantages of the
model. (10)
(c) Draw the activity diagram of project planning process. (10)
(d) Describe the difference between plan-driven and agile approaches with a diagram.

What are the advantages and drawbacks of agile approaches?

Contd P/4

,
\
.. ,
,

-.--., :1-
 •
;'

•
•

=4=

CSE 307

8. (a) The following table sets out a number of tasks of a project, their durations and
dependencies. How many days are required to complete the project? Draw a bar chart

showing the project schedule.

Task Duration (days) Dependencies

TI 10
T2 15 TI
T3 10
T4 20 T2, T3
T5 15
T6 10 T3,T5
T7 20 T3
,
T8 25 T6
T9 15 T4
TIO 5 T5, T9
TIl 10 TI, T3, T9
TI2 20 TIl

(b) Suppose you have got a project to develop a fully functional system within a very short
time period. Which process model will you choose to complete such a project? Illustrate

thc model. What are the factors that may influcnce you to reject the project offer? (15)
(c) Briefly discuss about the risk management process with necessary diagram. Write

down the potential indicators for each of the following risk type: (9+6=15) .
(i) People
(ii) Organizational
(iii) Estimation
.'

L-3ff-1ICSE Date: 18/0112017

BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY, DHAKA
L-3/T-l B. Sc. Engineering Examinations 2015-2016
Sub: CSE 305 (Computer Architecture)
Full Marks: 210 Time: 3 Hours
The figures in the margin indicate full marks.
USE SEPARATE SCRIPTS FOR EACH SECTION

SECTION -A
There are FOUR questions in this section. Answer any THREE.

1. (a) Consider three different processo,rs PI, P2 and P3 executing the same instruction set.
PI has a 3 GHz clock rate and a CPI of 1.5. P2 has a 2.5 GHz clock rate and a CPI of 1.0.
P3 has a 4.0 GHz clock rate and has a CPI of 2.2. (10)
(i) Which processor has the highest perfommnce expressed in instructions per second?
(ii) If the processors each execute a program in 10 seconds, find the number of cycles
and the number of instructions.
(iii) We are trying to reduce the execution time by 30% but this leads to an .increase
of 20% in the CPI. What clock rate should we have to get this time reduction?
(b) The following table shows the numbers of instructions for a program. (10)
Arithmetic Store Load Branch
500 50 100 50

The following table shows the number of cycles taken by each instruction.
Arithmetic Store Load Branch
1 5 5 2

(i) What is the execution time of the program in a 4 GHz processor?

(ii) What is the CPI of the program?
(iii) If the number of load instructions can be reduced by half, what is the speed-up
and CPI?
(c) Assume for arithmetic, load/store, and branch instructions, a processor has CPls of 1,
12, and 5, respectively. Also, assume that on a single processor a program requires the
execution of 2.56 billion arithmetic instructions, 1.28 billion load/store instructions, and
256 million branch instructions. Assume that each processor has a 2 GHz clock
frequency. Assume that, as the program is parallelized to run over multiple cores, the
number of arithmetic and load/store instructions per processor is divided by 0.7 x p
(where p is the number of processors) but the number of branch instructions per processor
9 6
remains the same. (I billion = 10 , 1 million = 10 ). (15)
(i) Find the total execution time for this program on I, 2, and 4 processors, and show
the relative speedup of the 2, and 4 processors compared to the single processor.
.(i) If the CPI of the arithmetic instructions was doubled, what would the impact be on
the execution time of the program on 1,2, and 4 processors?
(iii) How much should the CPI of load/store instructions be reduced for a single
processor to match the performance of 4 processors using the original CPI values?

Contd 1'/2
 =2=
CSE 305

2. (a) Consider the following information for two different computers. (5)
Computer Number of Instructions Clock Rate CPI
A 10 Billion 4 GHz I
B 8 Billion 4 GHz l.l

(i) Which has the highest MIPS rating?

(ii) Which is faster?
(b) Consider the following information for load word (lw) instruction (5)
Instr. Instruction Register ALU op Memory Register
Fetch read access write
lw 200 ps 100 ps 200 ps 200 ps 100 ps

Consider the following program:

lw$l,IOO($O)
lw $2, 200($0)
lw $3,300($0)
What is the speedup for the pipelined dadapath over the single cycle datapath for this
program? What will be the speedup if we add 2,000,000 more load word (lw) instructions?
(c) Draw the complete MIPS pipelined datapath and control for load (lw) instruction with
pipelined registers. (10)
(d) Consider the following C code (10)
A = B + C; D = A + E;
All the above are memory instructions. The base address is in register to and the offset
are A(O), B(4), C(8), D(l2), E(l6). How many cycles are required to execute the above?
Can you reduce the number of cycles using code scheduling? If yes, what will be the
resulting number of cycles? Show each step.
(e) "Even with a perfect branch predictor, I-cycle penalty for a taken branch". Why is
that? How to solve this? (5)

3. (a) What do you mean by double data hazard? Write down the equations to detect
forwarding in MIPS considering double data hazard? How MIPS detect hazard due to the
use ofload instruction? (10)
(b) Briefly explain write-through and write-back scheme for cache? How does write-
though and write-back schemes handle write-miss? (10)
(c) Briefly explain how page table works with respect to virtual address, physical address,
main memory and disk. (5)
(d) Though both TLB and data cache are cache memory, how are they different from
each other? Explain with flow chart the interaction between TLB and data cache for both
read and write operation. (10)
Contd P/3
.'

=3=
CSE 305

4. (a) Assume the miss rate ofan instruction cache is 3% and the miss rate of the data cache
is 5%. If a processor has a CPI of 2 without any memory stalls and the miss penalty is
100 cycles for all misses, determine how much faster a processor would run with a
perfect cache that never missed. Assume the frequency of all loads and stores is 40%.
What is the amount of execution time spent on memory stalls? (5)
(b) For a direct-mapped cache design with a 32-bit address, the following bits of the
address are used to access the cache. (1 0)
Tag Index Offset
31-10 9-5 4-0

What is the cache block size (in words) and how many entries does the cache have?
Starting from power on, the following byte-addressed cache references are recorded.
0,4,16,132,232,160,1024,30,140,3100,180,2180
How many blocks are replaced and what is the hit ratio?
(c) Assume there are three small caches, each consisting if four one-word blocks and uses
LRU replacement policy. One cache is fully associative, a second is two-way set-
associative, and the third is direct-mapped. Find the number of rilisses for each cache
organization given the following two independent sequences of block addresses: (15)
0,2,4,8,10,12,14,16,0
1,3,5,1,3,1,3,5,3
(d) Suppose we have a processor with a base CPI of 1.0, assuming all references hit in the
primary cache, and a clock rate of 4 GHz. Assume a main memory access time of 100 ns,
including all the miss handling. Suppose the miss rate per instruction at the primary cache
is 3%. How much faster will the processor be if we add a secondary cache that has a 5 ns
access time for either a hit or a miss and is large enough to reduce the miss rate to malO
memory to 1%. (5)

SECTION -B
There are FOUR questions in this section. Answer any THREE.

5. (a) Produce the compiled MIPS code for the following C function. Assume that addresses
ofdest, source in SaO, and $al. Use a minimal number of MIPS assembly instructions. (25)
void rstrcopy (char *dest, char *source)

if (*source)

*dest++=*source++;
rcopy (dest, source)

else
*dest='\O';

Contd P/4
•
=4=
CSE 305
Coutd •.. Q. NO.5

(b) What is the MIPS assembly code to load this 32-bit constant into register $sO? (6)
0001 0000 0011 1101 0000 1001 0000 0000
(c) What do you understand by pseudo instructions? Give two examples of
pseudo instructions that MIPS assembler accepts. (4)

6. (a) What are the steps of floating point addition? Describe with an example. (12)
(b) What are the different addressing modes of MIPS? Describe and illustrate the
calculation process for each addressing mode. (15)
(c) What is data race? What instructions are available in MIPS to address this issue? How
these instructions work? (8)

7. (a) Assume that you have four different threads. Show how the four threads could be
combined to execute on the processor more efficiently using following three
multithreading options: (15)
(i) A superscalar with coarse-grained l11Ultithreading
(ii) A superscalar with fine-grained multithreading
(iii) A superscalar with simultaneous multithreading
How the four threads would execute independently on a superscalar without any
multithreading support is given below.

I
: Issue Slots ~

Thread A Thread B Thread C Thread D

A A B B B C C C D
A B B I
Time A A A B D D D
A A B C C D D
A A A A
j B

B B B BI
C
C
C C C
D
D
D
D D D

B B
A B B
A A A I
I

(b) Find the maxImum number of processors each having clock rate i with cache
memories that a bus with Bandwidth B can support. (10)
(c) What is the use of Bias in floating point representation? Give the encoding of floating
point infinity. (10)
Contd P/5
•

=5=

CSE 305

8. (a) What are the different methods to maintain coherence among all caches and global

memory in a shared mcmory system? Describe briefly. (10)

(b) Suppose we want to sum 64,000 numbers on a shared memory multiprocessor
computer with uniform memory access time. Let's assume we have 64 (i) UMA

processors, (ii) message passing processors. Write the corresponding C code. (20)
(a) for each processor to sum their subset of numbers
(b) to add the partial sums

(c) Draw the organization of a message passing multiprocessor. (5)

...
L-4rr -2/CSE Date: 29/0112017
BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY, DHAKA
L-4/T-2 B. Sc. Engineering Examinations 2014-2015

Sub: CSE 311 (Data Communication)

Full Marks: 210 Time: 3 Hours

The figures in the margin indicate full marks.
USE SEPARATE SCRIPTS FOR EACH SECTION

SECTION-A
There are FOUR questions in this section. Answer any THREE.
Give your answer to the point avoiding unnecessary writings.
Try to draw as many figures as possible and explain with block diagrams.

1. (a) Briefly explain the steps required for "Echo Cancellation"? Explain how "Inter-

Symbol Interference (ISI)" can occur? (7+6)

(b) Briefly explain the procedure of 4B/5B encoding scheme and intentions behind it. (10)
(c) Show with necessary figures how 4-bit information is passed at a time with 16-

QAM (Quadrature Amplitude Modulation). (6)

(d) Fora baseband signal of m(t) = cos lOOt + cos 400t (6)
(i) Draw basic spectrum of m(t)
(ii) Draw spectrum ofthe modulated signal using Double Side Band with Carrier

(DSB+C), where the carrier signal is 2(1+ m(t ))cos 1OOOt. (6)

2. (a) Draw digital signals of the corresponding binary sequence of 1011001 for the

following line encoding schemes, (9+6)

(i) Pseudo-Ternary
(ii) Differential Manchester
(iii) Multi-transition (MLT)-3
Summarize each of these in terms of Bandwidth, Baseline wandering, and DC
component issues.
(b) Explain the incoherent demodulation procedure of FM and FSK. Explain how
modulations of Phase Shift Keying (PSK) and Quadratic Amplitude Modulation

(QAM) relate to each other? (5+3)

(c) Explain how offset QPSK (Quadrature Phase Shift keying) reduces maximum

phase shift from it to ~ without reducing bit rate. Explain how phase continuation is

maintained during MSK (Minimum shift keying). (6+6)

Contd P/2
.-

=2=

CSE 311

3. (a) Show that, for a real time f(t), the signal fUSB(t)= f{t)coswJ- fh (t)Siil wet
represents an upper sideband SSB-SC signal, where fh (t) represents Hilbert Transform

of f(t) and we is the carrier frequency in rad/sec. (13)

(b) What is orthogonal signal? Explain the motivations behind making each m-FSK
signals orthogonal to each other? Show that the minimum frequency difference to

make signal orthogonal is, Iif = 2~ Hz. (2+4+6)

b

(c)What was the fallacy behind the estimation of the bandwidth of Frequency
Modulation (FM)? Show that the bandwidth requirement of Wide Band Frequency
Modulation (WBFM) is '" 2(!::J.f + 2B) Hz , where !::if refers to peak frequency
deviation. (4+6)

4. (a) Explain with diagram how Phase Locked Loop (PLL) maintains Phase and
Frequency synchronization between source and generated carrier signals. (15)
(b) For a Vestigial Side Band (VSB) modulation with suppressed carrier, derive
,
following relation between input and output filter, H 0 - H ( ) I H (_ ). (8)
i CO + roc + i ro roc
(c) Explain with necessary figures and diagrams, the modulation and demodulation
techniques of On Off Keying (OaK), Frequency Shift Keying (FSK), and Differential
Phase Shift Keying (DPSK). (4+4+4)

SECTION-B
There are FOUR questions in this section. Answer any THREE.

5. (a) There are two principal forms of digital switching in communication networks, (10)
(i) Circuit switching
(ii) Packet switching
Big website companies like Facebook and Google store their \lata in data centers where
thousands of computers act together as a single server. Also, they establish multiple
data centers in geographically distance locations. That means when you load your
Facebook home page, it may be the case that the picture of Aashik Salam, one of your
computer science teachers, was downloaded from the Facebook data center at
Singapore and the picture of Himel Sen, another. of your computer science teachers,
was downloaded from the data center at California. However, it is not just the case that
only clients communicate with the data center servers, data centers communicate with
each other a lot as well. What form oLdigital switching do these data centers use to
communicate with each other?
(b) State and prove the dilation property of the Fourier transform. (5+5)
(c) If the Fourier Transform of the signal g(t) is G(j), then determine the Fourier

Transform ofthe signal, sin2(21tfct )g(t). (10)

(d) Derive the Fourier Transform of the Dirac Delta Function. (5)
Contd P/3
 =3=

CSE 311

6. (a) Determine the bandwidth of the following two signals. The frequency domain
representation of the signals are drawn as follows. Also tell what will be the bandwidth
if the diagrams drawn are for time domain representation. (3+3+4)
(i) Signal One .

-w w

(ii) Signal Two

-w, w -w F

(b) Derivation of the Fourier Transform of periodic signals are given below. Read the
derivation given below and write answer the questions: (20)
Let gr,(t) be a periodic signal with period To' The diagram and the definition of gr,(t)

regarding complex exponential Fourier Series are given below,

9r (t)-
'/
..'

I'
I
I

""
gro(t) = 2>n expv27t1ifot)
.
(I)
n=-oo

where Cn is defined as

cn = io
~ gr,(t )exp(- j27t1ifot) dt : (2)

Now, a pulse like function g(t) is defined in terms of gr.,(t) as,

g(t)={gr,(t) ; - ~ S;t S; ~ .....••....................... (3)

o ; elsewhere
Contd P/4
 =4=

CSE311
Contd ... Q. No. 6(b)

(i) Why do we need to introduce this new function g(I)? [ Given: the Fourier

Transform of g(t) if G(j)]

(ii) Also, Draw the time domain diagram of g(I).

Now en can be written as: cn ~ 10G(nlo) (4)

(iii) How can cn be equal to 10G(n10) ?

Now gr, (t) can be expressed as:

•
'"
gT.(I) = L:loG(nlo)exp(j27inlol) (5)
n=--oo

and the Fourier Transform of gT, (I) would be,

F [gT.(1)]= I/oG('?!0)F[exp(j21lnfot)] (6)

n=-aJ

(iv) Why is Fourier Transform only applied to exp(j21l'?!01)?

Why is Fourier Transform not applied to the part 10G(nlo)?

From this definition we can conclude,

F[gT. (1)]= 10 IG(nlo)8(j -nlo) (7)

n=-oo

(v) How can we deduce Eq. 7 from Eq. 6?

(c) Why is non-uniform quantization needed? What are the problems III case of

uniform quantization? (5)

7. (a) Write short notes on- (5+5)

. (i) Linear System
(ii) Time Invariant System
(b) Draw the response of the ideal low-pass filter if the following pulse was given as
input. From these two diagrams, explain why this filter is called "ideal" rather than

"real"? (10)

,.

(c) Deduce the recurrence relationship for the Fast Fourier Transform. (15)
Contd PIS
 =5=

CSE 311

8. (a) Suppose a Fourier Transform pair signai is given as follows: (6+3+6)

get) G(t)

-.~---.

The signal is sampled as described in the following figure,

get)

The sampled signal is described as,

g&(t)=g(t) L'" o(t-nTJ= L'" g(nTJo(t-nTJ

n=-«l n=-oo

Find the Fourier Transform of g&(t). Also draw the diagram of G&V). From the

(10)

(10)
..

L-3/T-lICSE Date: 04/02/2017

BANGLADESH UNIVERSITY OF ENGINEERING AND TECHNOLOGY, DHAKA
L-3/T- I B. Sc. Engineering Examinations 201 5-201 6

Sub: CSE 303 (Database)

Full Marks: 210 Time: 3 Hours

The figures in the margin indicate full marks.
USE SEPARATE SCRIPTS FOR EACH SECTION

SECTION-A
There are FOUR questions in this section. Answer any THREE.

I. (a) A distributed database system consisting of five sites is shown in Figure I. A

transaction, TJ (Figure 2) has been initiated from site SJ. Data item A is fragmented
and stored in sites S2 and S4. Data item B is fragmented and stored in sites S2, 83 and

S5. Accordingly, any update (write) operation need to be performed to all related sites. (20)
(i) Show all the messages e.g. <Prepare TJ> etc. passing between SJ (transaction
initiator) and the related other sites (S2, S3, S4 and S5) for phase 1 and phase
2 to COMM1T the transaction TJ using the two-phase commit protocol.
(ii) Show all the message passing similar to I (a)(i) for phase I and phase 2 to
abort the transaction T1 because of sending < Abort TJ> message from
S5 to SJ.

(b) Figure 3 shows the fixed length record file structure for customer relation. Show
the file structure after deletion of records with customer ids C004 and C007 for
different file deletion methods. Compare the advantages and disadvantages among

different deletion methods as per above. (15)

2. (a) Find the equivalent serial schedule for the concurrent schedule shown in Figure 4 by

swapping of non-conflicting instructions. (15)

(b) Verify the serial schedule created in 2(a) with the serial schedule created by using

precedence graph. (5)

(c) Create lock table for the list of data items and the corresponding locking status of

different transactions as given in Figure 5. (15)

3. (a) A set of transactions along with their time stamp values and instructions is given in
Figure 6(b). The timestamp values of data item A is given in Figure 6(a). Find the
status (roll back or executed) of each transaction and the corresponding timestamp

values of data item A using the timestamp-based protocol. (15)

(b) Explain Thomas write rule with an example. (5)

Contd P/2
 =2=

CSE 303
Could ... O. No.3
(c) (i) The initial account balance of A), A2 and A3 are 1000, 2000 and 3000

respectively. Write log records for the transactions given in Figure 7. (15)
(ii) Show the log-based recovery if a failure occurs after write (A3).
(iii) Show the recovery if a failure occurs after write (A2).

4. (a) Construct the dynamic hash index structure on balance for the relation given in
Figure 3. The binary representation of the most significant digit of balance is the hash
code on balance. You have to insert records in descending order of balance. (20)
(b) (i) Show the final B+ tree index structure for the search key values 2, 4, 6, 8,
10,12,14,16,18,20,22,24 with 4 leaf nodes and n = 4. You do not need to
create this index in step by step. (10)
(ii) Show the index structure after insertion of 15.
(c) Write short note on bitmap indexing. (5)

SECTION-B
There are FOUR questions in this section. Answer any THREE.
Please read carefully, some questions might have additional restrictions.

5. (a) Consider the following relational database schema: (7x4=28)

Product (maker, model, type)
PC (model, speed, ram, hd, price)
Laptop (model, speed, ram, hd, screen, price)
Printer (model, color, type, price)
The Product relation gives the manufacturer, model number and type (PC, laptop, or
printer) of various products. We assume for convenience that model numbers are
unique over all manufacturers and product types; that assumption is not realistic, and a
real database would include a code for the manufacturer as part of the model number.
The PC relation gives for each model number that is a PC the speed (of the processor,
in gigahertz), the amount of RAM (in megabytes), the size of the hard disk (in
gigabytes), and the price. The Laptop relation is similar, except that the screen size (in
inches) is also included. The Printer relation records for each printer model whether
the printer produces color output ('T', if so; 'F' otherwise), the process type (laser or
ink-jet, typically), and the price. Write the following SQL queries:
(i) Find those manufacturers that sell Laptops, but not PCs or Printers.
(ii) Find those hard disk sizes that occur in two or more PCs.
(iii) Find those pairs of PC models that have both the same speed and RAM. A
pair should be listed only once.
(iv) 'Find the manufacturers of PCs with at least three different speeds.
(v) Find the makers who produce only one type of product.

Contd P/3
 =3=
CSE303
Contd ... O. No. Sea)
(vi) Find the maker(s) of the cheapest color printer(s).
(vii) Find the PC model with maximum processing speed without using any
aggregate function.
(b) Consider three tables Tl, T2 and 13. Each table has only one column which
contains only integers. Suppose the content of the three tables are as follows: (2+2+3=7)
T1: 1,2,2,2,3,4,4
T2: 2,3,4,4,4,5
T3: 2,2,3,3,4,4,5, 5
Show the values ofthe following SQL queries:
(i) SELECT * FROM Tl UNION SELECT * FROM T2
(ii) SELECT * FROM Tl MINUS SELECT * FROM T3
(iii) (SELECT * FROM Tl UNION ALL SELECT * FROM T2) INTERSECT
ALL (SELECT * FROM T3)

6. (a) Consider a relation R(A, B, C, D) and FD's AB~C, BC~D, CD~A, and AD~B. (12)
(i) What are all nontrivial FD's that follow from the given FD's? You should
restrict yourself to FD's with single attributes on the right side.
(ii) What are all the keys ofR?
(iii) What are all the superkeys for R that are not the keys?
(b) Prove that any two-attribute relation is in BCNF. (7)
(c) For the relation schema R(A, B, C, D, E) with FD's AB~C, DE~C, and B~D do
the following: (10)
(i) Indicate all the BCNF violations.
(ii) Decompose the relations, as necessary, into the collections of relations that
are in BCNF.
(d) Suppose R is a relation with attributes AI, A2, A3, ..., An. As a function ofn, find
out how many superkeys R has, if the only keys are {A l} and {A2, A3}. (6)

7. (a) Consider the following relational database schema designed for a university
database: (4x4=16)
Department (name, headID, yearOfEastablishment)
Employee ili!, name, address, designation, salary, gender)
(i) Suppose we want to enforce the rule that says "Departmental head must be
an employee ofthe university". Show two ways to implement the above rule:
one using a check constraint and the other without using any check
constraint.
(ii) Write a check constraint to make sure that no two departments have the same
person as Head.
(iii) Write an assertion to check the rule: if the Head of the Department's gender
is male then his name must not begin with 'Ms'.
(iv) Write an assertion to check the rule: the CSE Departmental head's salQlY must
not be less than 1,50,OOOTaka.

Contd P/4
 =4=
CSE 303
Contd ... Q. NO.7

(b) (i) Consider the following DOL: (4+4=8)

CREA TE TABLE MyTabIe(

gender CHAR(l) NOT NULL,

);

Replace the above NOT NULL constraint using a CHECK constraint.

(ii) Consider the following DOL:

CREATE TABLE MyTabIe(

headID INT UNIQUE,

);

Replace the above UNIQUE constraint using a CHECK constraint.

(c) What are the three different policies for UPDATE and DELETE that you can
mention for maintaining referential integrity? Among these which one is more suitable
for UPDATE and which one is more suitable for DELETE? Why? (6)
(d) Suppose we want to create a copy of a table R including the schema. Show the
DOL to perform such operation. What will be the DOL if we just want to copy the
schema and not the tuples of the table R? (5)

8. (a) East West Property Development is a construction company of Bashundhara group

with over 1000 employees. A customer can hire the company for more than one
project, and employees sometimes work on more than one project at a time. Equipment
is assigned to only one project. Draw an E-R diagram for this company indicating
cardinality. (7)
(b) A university has a large number of courses. Each course may have one or more
different courses as prerequisites, or may have no prerequisites. Similarly a particular
course may be prerequisite of any number of courses or may not be prerequisite of any
other courses. Sketch an E-R diagram that includes COURSE as the single entity,
including cardinality. Each course has a course number and title. Convert the E-R
diagram into relation schema. (7)

Contd PIS
 =5=
CSE 303
Contd ... Q. NO.8

(c) Convert the following E-R diagram into Relation schema: (6)

"

Company
Product

Person

Figure 8: E-R diagram for question 8(c)

(d) Consider a database with the following schema (4+4+4+3=15)

Route: flight number (FN), departure city (DC), arrival city (AC), distance in miles
(D!), scheduled department time (SD), Scheduled arrival time (Sa), plane model (PM).
Flight: flight number (FN), date (DT), number of passengers (NP), actual department
time (AD), actual arrival time (AA),
Plane: plane model (PM), capacity (CA), type (TP).
A snapshot of the database is shown in Figure 9. Translate the following queries into
relational algebra:
(i) Which flights have at least 50 passengers on board arrived more than 30
minutes late?
(ii) What are the departure and arrival cities of all flights on October 10th that
were completely full?
(iii) What are the plane types that travel between BOSTON (BOS) and
FRANCISCO (SFO)?
(iv) Which flights departed late but arrived without any delay?

Contd P/6
 =6=
CSE 303
Contd ... Q. No. Sed)

Figure 9: A sample snapshot of database for question 8Cd)

p.T.O
 -1--

Figure 1: Distributed Database 5ystem consisting of five sites

Data replication
T1
READ (A)
Data item A is fragmented to 52, 54
READ (6)
WRITE (A) Data item B is fragmented to 52, 53, 5S
WRITE (6)
• COMMIT

Figure 2: A transaction T1 initiated from 51 of Figure 1.

Name street City Balance

Customer Id
N1 North Dhaka 1000
record 0 COOOl
N2 North Dhaka 2000
record 1 COO02
N3 North Dhaka 3000
record 2 COO03
N4 North Dhaka 4000
record 3 COO04
N5 North Dhaka 5000
record 4 C0005
N6 South Dhaka 6000
record 5 C0006
N7 South Dhaka 7000
record 6 COO07
N8 South Dhaka 8000
record 7 COO08
N9 South Dhaka 9000
record 8 COO09
N10 South Dhaka 10000
record 9 C0010.

Figure 3: File containing Customer relation records _~ __ -J

p.1.o

.".,
•

7 Tl
READ(A)
T2

READ (e)
T3 T4
I
,
,

READ(S) . ,
I
WRITE (e) i
READ(e)
WRITE (e)
REAR(S)
READ(A)
READ(D)
WRITE (A)
WRITE (D)
READ(S)
i
i
READ(S) II
i
:
Figure 4: Concurrent schedule of four transactions

Data item Hash Value Transactions Transactions waiti!'lg

..
obtained lock for lock
A 5 Tl,T5 T2
S 3 Tl,T2
D 5 T3
F 2 T2 T5
G 2 T2, T3, T4 Tl
H 5 T4

Figure 5: List of data items, corresponding hash values and transactions obtained lock
and waiting for lock.

p. r. 0
 •
•

Data Item time Time stamp value

. stamp
W-timestamp (A) 10
R-timestamp (A) 20
(a)

Transaction Timestamp Instruction

..
T1 5 READ (A)
. T2 15 READ (A)
13 25 WRITE (A)
T4 12 WRITE (A)
(b)

Figure 6: (a) Timestamp values of data item A

(b) Timestamp values of transactions

T20 .. T21 .

READ (A1)
A1=A1+100
WRITE (A1)
READ (A3)
A3 =A3-100
WRITE (A3)
COMMIT
READ (A2)
A2=A2+0.1*A2
WRITE (A2)
'C
COMMIT

Figure 7: Two serial transactions

/.

--.-