Research Methodology EPH 7112 LECTURE 6: RESEARCH PLANNING.

Research Methodology

EPH 7112

LECTURE 6: RESEARCH PLANNING

THE ISSUES

Sometime students are not clear on: 1. How to start 2. What area, what topic 3. How to choose a supervisor 4. What’s the scope 5. How to get data 6. How to analyze

7. When to stop

MATERIALS TO COVER:

PART A: THE PHILOSOPHY

PART B: THE PLANNING

PART C: THE IMPLEMENTATION

Note: the part on “What Your Supervisor Expects from You” is not covered here

THE WHOLE PROCESS OF RESEARCH STUDY

IDENTIFY TOPIC

OBTAINED DATA

IDENTIFY RESEARCH AREA

START

LITERATURE REVIEW

CHOOSE THE SUPERVISOR

DEVELOP STUDY PLAN/MODEL

ANALYZED RESULT

OUTPUT /CONCLUSION

THESIS SUBMISSION

VIVA

THESIS WRITING

EXECUTE RESEARCH

LITERATURE REVIEW

PRODUCING RESULTS

OUTPUT/CONCLUSION

PERFORM ANALYSIS ON

RESULT

START

DATA

RESULT

END

YES

NO

PROCESS GENERATING & COLLECTING DATA

GRAPH, TABLE, CHART

Processes in obtaining data (Quantitative/Qualitative)

Theoretical modeling/Software simulationb. Emulationc. Survey d. Experimental

It’s all about Testing the system under study

Data normally come in the form of Tables

Results normally come in the form of Graphs Charts

Output/Conclusion (the system under study) Working system Product New Process Ingredients Algorithm Programs Formula

Analyzing Results

Results consist of : 1. Presentation of data 2. Analysis Analysis of results must include the explanation on:Design Parameters (Cause) and Performance Parameters (Effects) of the System Under Study.

A comparative analysis is typically required.

SYSTEM UNDER STUDY

DESIGN PARAMETERS

PERFORMANCE PARAMETERS

Output/Conclusion

Conclusion:Targets/Objectives achieved, how much?Which results directly support the conclusion?

Output can be :- 1. Formula 2. Process 3. Algorithm/Program 4. Prototype 5. System

Novelty

1. Theory

2. Process/techniques/fabrications

3. Ingredients/Materials

4. Design/architecture

5. Properties/Characteristics

PART B: THE PLANNING

Study Plan

A Study Plan is a document explaining the plan of the study/research project from its start to the end. It is normally presented in a Gantt chart, and it should clearly show:

1.The research activities2.The continuity of activities3.The period of activities/timeline 4.Deliverables and milestones5.The end goal

Study Plan

It basically addresses the issues of:1. What to do

2. When to do

3. What to achieve

4. When to stop

Deliverables and Milestones

1.All outputs are Deliverables

2.Major outputs are Milestones

A milestone represents an important stage of a project.

A sample study plan

YEAR 2002 2003 2004

MONTH 2 3 4 5 6 7 8 910

11

12 1 2 3 4 5 6 7 8 9

10

11

12 1 2 3 4 5 6

No. Activity                                                          

1Literature Review                                                          

2Code Development            

*                                            

3Computer Simulation                                                          

4 Procurements.                                              *           

5 Experimentation                                                          

6 Result Analysis                                                      *   

7 Report Writing                                                          

Weaknesses of a Study Plan

1. It does not clearly show the structure/scope; depth and width of the project

2. It does not provide the designation of issues under study; general issues,

complementary issues and focused issues

3. It does not adequately indicate the number of results expected of a project

Study Model (K-Chart)

A Study Model is a document explaining the structure, issues and sub-issues under study. It is usually presented in the form of a Tree Diagram. A K-Chart should clearly show:

1. The issues of concern 2. The sub-issues of concern 3. The designation of theory, simulation and experiment 4. The Design and Performance parameters 5. The line of focus 6. The scope (depth and width) of a project

K-Chart should be able to give an idea on:

1. How many results will be obtained.

2. Which results are to be highlighted.

3. How many chapters (and what chapters) will there be.

4. The coverage of the literature review.

5. How many pages the thesis will be.

How to construct a K-Chart

1. Identify the area of study (Main title)2. Identify the main issues under the area of study – these issues

come as branches under the Title3. Identify the sub-issues under the main issues – again as lower

layer branches4. End the issues/sub-issues once you get the required info for you

to decide (which one you want to proceed into the details in)5. Continue step 3 until you reach the data generating methods

(experiment, simulation, survey etc)6. End the structure with a list of Design (DP) and Performance

Parameters (PP) 7. You may have redundant issues and/or redundant DPs/PPs – not

to worry! Whichever way you do, you should end with same results

8. Highlight the tree branches that you want to study/focus

SAMPLE STUDY MODEL:OPTICAL MULTIPLEXING TECHNIQUE

MULTIPLEXING TECHNIQUE

OCDMA FDMA TDMA

ADVANTAGES DISADVANTAGES DISADVANTAGES

Sample Study Model(CONSTRUCTION AND IMPLEMENTAION OF A NEW CODE DIVISION MULTIPLEXING TECHNIQUE FOR METROPOLITAN AND ACCESS NETWORK)

2

OSCDMA

CODE DEVELOPMENT

APPLICATIONS DEVICES

MAN LAN ACCESS ENCODER DECODEROTHER CODES

MAIN ISSUES

SUB ISSUES

DW MDW MUX/ DEMUX

OS-CDMA APPLICATION

MAN ACCESS

SYSTEM DESIGN(POINT TO POINT )

SYSTEM DESIGN ( POINT

TO MULTIPOINT )

MODULATION TECHNIQUE

SOURCE

SUBTRACTION

TECHNIQUE

DW MDW

MODULATION

TECHNIQUE

SOURCE

SUBTRACTION

TECHNIQUE

APPLICATIONS

MODULATION TECHNIQUE

SOURCE

SUBTRACTION

TECHNIQUE

DW

MDW

MODULATION TECHNIQUE

SOURCE

SUBTRACTION TECHNIQUE

BIT RATEBIT RATE

BIT RATE

BIT RATE

MAIN ISSUE

SUB ISSUES

Application – MAN (Simulation)

MAN

SYSTEM DESIGN(POINT TO POINT)

MDW

MODULATION TECHNIQUE SOURCE

SUBTRACTION TECHNIQUE

DW

EXTERNAL SINGLE MULTIPLE COMPLEMENT ANDDIRECT

BIT RATE

SAME BIT RATE

VARIABLE BIT RATE

SOURCESUBTRACTION

TECHNIQUE

SINGLE MULTIPLE COMPLEMENT AND

BIT RATE

SAME BIT RATE

VARIABLE BIT RATE

D.P = DESIGN PARAMETERPP = PERFORMANCE PARAMETER

DistanceInput PowerChip spacingBit Rate

D.P

BERQ FactorLoss(Prx)

P.P

WEIGHT

SINGLE MULTIPLE

MAN

SYSTEM DESIGN( POINT TO POINT)

MODULATION TECHNIQUE SOURCE

SUBTRACTION TECHNIQUE

DW

EXTERNAL SINGLE MULTIPLE COMPLEMENT ANDDIRECT

BIT RATE

SAME BIT RATE

VARIABLE BIT RATE

DistanceInput Power

D.P

BER

Loss ( Prx )

P.P

D. P = DESIGN PARAMETERPP = PERFORMANCE PARAMETER

Application – MAN (Experiment)

SYSTEM UNDER STUDY(Metropolitan Area Network- Point To Point)

1. SYSTEM 1 ( DIRECT MODULATION OSCDMA ) 2. SYSTEM 2 ( EXTERNAL MODULATED OSCDMA)3. SYSTEM 3 (MULTIPLE SOURCE OSCDMA)4. SYSTEM 4 (SINGLE SOURCE OSCDMA)5. SYSTEM 5 (COMPLEMENT SUBTRACTION)6. SYSTEM 6 (AND SUBTRACTION)7. SYSTEM 7 (SYNCHRONOUS)8. SYSTEM 8 (ASYNCHRONOUS)

DESIGN PARAMETER vs PERFORMANCE PARAMETER

Design Parameters: Performance Parameters

1. Distance (Fiber) 1. BER (Bit error rate)

2. Input Power (PTX) 2. Q Factor

3. Chip Width 3. Power Receive (PRX)

4. Bit Rate  

WHAT STUDY MODELS GIVE US

1. A clear title of project/thesis

2. The scope, and objective (abstract)

3. The un-highlighted issues – Literature reviews (Background reviews, Critical reviews, Problem Statement)

4. Number of results

How to Calculate the Number of Results

No Of Results = No of system x No of Design Parameters x No of Performance Parameters

Example : Total Number Of Results(OSCDMA Project: Metropolitan Area Network Point to Point)

Total Result = 8 (Number Of system) x 4 (Design Parameter) x 3 (Performance Parameter) = 96

PART C: THE IMPLEMENTATIONLITERATURE REVIEWS AND RESULTS ANALYSIS

LITERATURE REVIEWS

Consists of:

1. Background Review (Introduction)

2. Critical Review (throughout the thesis, typically in Problem Statement, conclusions)

LITERATURE REVIEWS

1. Start with text books, white papers, “review’ papers – background review

2. Then, with journal papers (always read more than once, at different times)

3. Review until you have enough info to decide – focus on Advantages & Disadvantages

4. How many references; > 50 for MSc, >100 for PhD

5. Index your references

LITERATURE REVIEWS, con’t

6. Jot down info you get from the references – in your reference note book

7. Focus on strengths/weaknesses of papers

RESULTS PRESENTATION

Typically, engineering research results come in the form of graphs/curves

Graphs: Y vs. X

Y axis – Performance Parameter (Output)

X axis – Design Parameters (Input)

RESULTS PRESENTATION con’t

For Multiple DPs, always present only one at a time. Analyze them individually, then only put them together when necessary (i.e, better presentation)

Note: Many times, students end up producing (too) many results, look different but similar (only realized after going through the analysis). Happens when trying to put multiple DPs at the same time.

RESULTS ANALYSIS, con’t

There are two main elements of analysis:

1. The Trend, 2.The Reason

Trend Parameters: Max, Min, Average (Mean), Slope, Skewness, Mod,Median

Reason: WHY the trend?

Characterization curve (CC) Vs Optimization curve (OC)

CC : Y Vs X1 at Fixed X2

OC : X1 Vs X2 at Fixed Y

THANK YOU