How To Exam?

a knowledge trading engine...


Birla Institute of Technology (BIT Mesra) 2007 B.E Electrical and Electronics Engineering Network Emdded Application - Question Paper

Saturday, 19 January 2013 02:25Web

Birla Institute of Technology & Science, Pilani
I Semester 2007-2008
EEEG627 Network Embedded Application
Comprehensive exam Part- A (Closed Book)
Date: 12-12-2007 Max Marks: 25 Duration: one Hr

Q1. What is the difference ranging from predictability in a dynamic real-time system and a static real-time system? (1)
Q2. What is a specification violation in RTS? (1)
Q3. “Real-Time programming is just optimal assembly coding, priority interrupt programming and writing device drivers”. actual or False? Elucidate. (1)
Q4. What is the function of the communication system in a networked Multimedia application? (1)
Q5. In A multi-processor system with m (application) processors, suppose the ratio of true to worst-case calculation time of all tasks scheduled on n processors, where n < m is 1.0 Under this situation will basic algorithm ever reclaim. Elucidate. (2)
Q6. When a task is parallelized, all parallel subtasks have to begin at the identical time. Devise a method to construct RVs for a resource-reclaiming algorithm in the presence of parallelized tasks in the feasible schedule. (3)
Q7. Construct a periodic task set with calculation time and periods that can be feasibly scheduled by EDF and not by RMS algorithm? (2)
Q8. What causes the true calculation time of a task to be less than its worst-case calculation time? (1)
Q9. A multi-processor system has 2 resources R1 and R2. All accesses to resource R1 are in shared mode, and those tasks that use R1 are scheduled on set, say S1 of processors. All accesses to resource R2 are in exclusive mode, and those tasks that use R2 are scheduled on the set, say S2, of processors. S1?S2 = ?. For what cardinalities of set S1 and S2 will run-time anomaly never occur if work-conserving algorithm is used for resource reclaiming? (2)
Q10. In an adaptive fault tolerant schedule, there is a flexibility of selecting the fault tolerant scheme, if the task permits o. If TMR and PB are the choices, when should a designer select TMR and when should he select PB? (2)
Q11. Does a system with high availability mean high reliability and vice versa? (1)
Q12. Is imprecise calculation a forward or backward recovery scheme? Justify. (2)
Q13. Was the MARS - Pathfinder RTS a networked embedded system or multi-processor system? Elucidate. What exactly was the issue that occurred with the pathfinder system? How was it solved? (2)
Q14. Is it possible that the performance of VTCSMA and PBCSMA protocols can fall as the message laxity increases beyond a certain point? discuss. (2)
Q15. What is delay-jitter in real-time communication? (2)

Birla Institute of Technology & Science, Pilani
I Semester 2007-2008
EEEG627 Network Embedded Application
Comprehensive exam Part- A (Closed Book)
Date: 12-12-2007 Max Marks: 55 Duration: two Hrs

Note: The ques. has 9 parts. As the issue progresses the ques. are discussed. If the 1st part is not answered correctly 2nd part will also be incorrect and so on, no marks are awarded for partial answers. If 1st part is incorrect then the entire ans will be incorrect. Rough work can be done in the additional sheets given. Only the final flow and the schedules must be shown on the main sheet.
Q1. Consider the network shown in fig 1







Fig1: Local area network of 21 nodes
System Details
every node is made of three processors. All tasks that arrive at every node have hard real-time constraints. The tasks are parallelizable. Ready Time of all tasks is zero.
Scheduling for Multi-processor System.
Scheduling is centralized. Scheduling is done using Par-Myopic /Myopic Algorithm. Parameters used by the scheduling algorithm are as follows: There is only a single instance of resource available. K –3, W-1, Number of back tracks = 0. Tasks 1, 8, 11 are multimedia tasks that require signal processing and hence can be executed on only processor one and 3.
Task Ti deadline di Worst case calculation time (ci1) Worst case calculation time (ci2) Actual exec time Resource Req
(R1)
1 100 50 50 25
2 250 175 100 75
3 350 75 60 50
4 500 200 150 100 share
5 600 200 150 100
6 150 150 125 125
7 200 50 50 50 Exclusive
8 300 75 75 75
9 600 75 75 75 Exclusive
10 50 50 50 50
11 300 100 80 50
12 450 200 150 100 share
13 575 175 120 100
What is the resulting schedule using the algorithm you have selected? Will Myopic algorithm be able to produce a feasible schedule? Elucidate. If not under what conditions will Myopic algorithm be able to produce a feasible schedule.
[Note: When 2 tasks have the identical deadline- task with lesser Id should be provided higher priority] [15]
Of the tasks in the Table - task one is safety critical. replace the scheduling algorithm accordingly. Are all the tasks schedulable? Show the resulting schedule. [6]
The true execution time of the tasks is provided in the Table. Based on the true execution-time what will be the post-run schedule? [3]
Resource-Reclaiming is done using Early-Start. What is the final schedule after resource reclaiming if all tasks are run successfully? [Note: If a task is parallelized then the true execution time is equal to the worst-case calculation time] [6]
Do you think the scheduling algorithm utilizes the processor efficiently? What could be the modification that can be made to the scheduling algorithm in order to increase the processor utilization capacity? [3]
info Policy
info policy followed makes use of the Maekawa set. Generate the Request set, Inform set and Status set for the nodes in the network. [6]
Location Policy
Location policy is sender initiated.
Message Scheduling
Messages are scheduled using CSMA-based DCR protocol.
Show the complete mapping on the UB Tree. [4]
When a message is being transmitted collision occurs and the colliding nodes are 2, 4, 6, 7, 8, 12, 14, 15, 16, and 20. Show the trend in which collision in resolved within an epoch. [6]
If DOD-CSMA is used and the message classes of the colliding messages are 3, 2, 3, 2, 1, 2, 3, 3, 2, and two for the nodes 2, 4, 6, 7, 8, 12, 14, 15, 16 and 20 respectively. What is the variation in the collision resolution pattern? [Note: F=3] [6]


Birla Institute of Technology & Science, Pilani

Birla Institute of Technology & Science, Pilani

I Semester 2007-2008

EEEG627 Network Embedded Application

Comprehensive Examination Part- A (Closed Book)

Date: 12-12-2007 Max Marks: 25 Duration: 1 Hr

 


Q1. What is the difference between predictability in a dynamic real-time system and a static real-time system? (1)

Q2. What is a specification violation in RTS? (1)

Q3. Real-Time programming is just optimal assembly coding, priority interrupt programming and writing device drivers. True or False? Elucidate. (1)

Q4. What is the function of the communication system in a networked Multimedia application? (1)

Q5. In A multi-processor system with m (application) processors, suppose the ratio of actual to worst-case computation time of all tasks scheduled on n processors, where n < m is 1.0 Under this situation will basic algorithm ever reclaim. Elucidate. (2)

Q6. When a task is parallelized, all parallel subtasks have to start at the same time. Devise a method to construct RVs for a resource-reclaiming algorithm in the presence of parallelized tasks in the feasible schedule. (3)

Q7. Construct a periodic task set with computation time and periods that can be feasibly scheduled by EDF and not by RMS algorithm? (2)

Q8. What causes the actual computation time of a task to be less than its worst-case computation time? (1)

Q9. A multi-processor system has two resources R1 and R2. All accesses to resource R1 are in shared mode, and those tasks that use R1 are scheduled on set, say S1 of processors. All accesses to resource R2 are in exclusive mode, and those tasks that use R2 are scheduled on the set, say S2, of processors. S1S2 = . For what cardinalities of set S1 and S2 will run-time anomaly never occur if work-conserving algorithm is used for resource reclaiming? (2)

Q10. In an adaptive fault tolerant schedule, there is a flexibility of selecting the fault tolerant scheme, if the task permits o. If TMR and PB are the choices, when should a designer choose TMR and when should he choose PB? (2)

Q11. Does a system with high availability mean high reliability and vice versa? (1)

Q12. Is imprecise computation a forward or backward recovery scheme? Justify. (2)

Q13. Was the MARS - Pathfinder RTS a networked embedded system or multi-processor system? Elucidate. What exactly was the problem that occurred with the pathfinder system? How was it solved? (2)

Q14. Is it possible that the performance of VTCSMA and PBCSMA protocols can fall as the message laxity increases beyond a certain point? Explain. (2)

Q15. What is delay-jitter in real-time communication? (2)


Birla Institute of Technology & Science, Pilani

I Semester 2007-2008

EEEG627 Network Embedded Application

Comprehensive Examination Part- A (Closed Book)

Date: 12-12-2007 Max Marks: 55 Duration: 2 Hrs

 


Note: The question has nine parts. As the problem progresses the questions are explained. If the first part is not answered correctly second part will also be incorrect and so on, no marks are awarded for partial answers. If first part is incorrect then the entire answer will be incorrect. Rough work can be done in the additional sheets provided. Only the final flow and the schedules must be shown on the main sheet.

Q1. Consider the network shown in fig 1

 

 

 

 

 

 

 


Fig1: Local area network of 21 nodes

System Details

Each node is made of 3 processors. All tasks that arrive at each node have hard real-time constraints. The tasks are parallelizable. Ready Time of all tasks is zero.

Scheduling for Multi-processor System.

Scheduling is centralized. Scheduling is done using Par-Myopic /Myopic Algorithm. Parameters used by the scheduling algorithm are as follows: There is only a single instance of resource available. K 3, W-1, Number of back tracks = 0. Tasks 1, 8, 11 are multimedia tasks that require signal processing and hence can be executed on only processor 1 and 3.

Task Ti

deadline di

Worst case computation time (ci1)

Worst case computation time (ci2)

Actual exec time

Resource Req

(R1)

1

100

50

50

25

 

2

250

175

100

75

 

3

350

75

60

50

 

4

500

200

150

100

share

5

600

200

150

100

 

6

150

150

125

125

 

7

200

50

50

50

Exclusive

8

300

75

75

75

 

9

600

75

75

75

Exclusive

10

50

50

50

50

 

11

300

100

80

50

 

12

450

200

150

100

share

13

575

175

120

100

 

What is the resultant schedule using the algorithm you have selected? Will Myopic algorithm be able to produce a feasible schedule? Elucidate. If not under what conditions will Myopic algorithm be able to produce a feasible schedule.

[Note: When two tasks have the same deadline- task with lesser Id should be given higher priority] [15]

Of the tasks in the Table - task 1 is safety critical. Modify the scheduling algorithm accordingly. Are all the tasks schedulable? Show the resultant schedule. [6]

The actual execution time of the tasks is given in the Table. Based on the actual execution-time what will be the post-run schedule? [3]

Resource-Reclaiming is done using Early-Start. What is the final schedule after resource reclaiming if all tasks are run successfully? [Note: If a task is parallelized then the actual execution time is equal to the worst-case computation time] [6]

Do you think the scheduling algorithm utilizes the processor efficiently? What could be the modification that can be made to the scheduling algorithm in order to increase the processor utilization capacity? [3]

Information Policy

Information policy followed makes use of the Maekawa set. Generate the Request set, Inform set and Status set for the nodes in the network. [6]

Location Policy

Location policy is sender initiated.

Message Scheduling

Messages are scheduled using CSMA-based DCR protocol.

Show the complete mapping on the UB Tree. [4]

When a message is being transmitted collision occurs and the colliding nodes are 2, 4, 6, 7, 8, 12, 14, 15, 16, and 20. Show the pattern in which collision in resolved within an epoch. [6]

If DOD-CSMA is used and the message classes of the colliding messages are 3, 2, 3, 2, 1, 2, 3, 3, 2, and 2 for the nodes 2, 4, 6, 7, 8, 12, 14, 15, 16 and 20 respectively. What is the variation in the collision resolution pattern? [Note: F=3] [6]


( 0 Votes )

Add comment


Security code
Refresh

Earning:   Approval pending.
You are here: PAPER Birla Institute of Technology (BIT Mesra) 2007 B.E Electrical and Electronics Engineering Network Emdded Application - Question Paper