Strategies for Providing Students and Researchers in ...
Strategies for Providing Students and Researchers in Developing Environments Access to Industry Standard Hardware and Software Technologies
by
Madhav Srimadh
Submitted to the System Design and Management Program
in Partial Fulfillment of the Requirements for the Degree of
Master of Science in Engineering and Management
at the
Massachusetts Institute of Technology
June 2003
© 2003 Madhav Srimadh
All rights reserved
The author hereby grants to MIT permission to reproduce and to
distribute publicly paper and electronic copies of this thesis document in whole or in part.
Signature of Author
Madhav Srimadh
System Design and Management Program
June 2003
Certified by
Amar Gupta
Thesis Supervisor
Co-Director, PROductivity From Information Technology
Accepted by
Steven D. Eppinger
Co-Director, LFM/SDM
GM LFM Professor of Management Science and Engineering Systems
Accepted by
Paul A. Lagace
Co-Director, LFM/SDM
Professor of Aeronautics & Astronautics and Engineering Systems
Executive Summary
Problem Statement
The author believes that students learn best when they practice the concepts learnt in the classroom through experiments and hands-on exercises. However, in many developing environments and some rural areas of developed environments, the academic institutions are simply unable to provide industry standard laboratories to their students. There are several factors contributing to this inability. First, given the strength of the local economies and weak policies surrounding indigenous manufacturing infrastructure, acquiring these technologies is an extremely expensive endeavor. Second, many of these improved technologies are not easily available. There are several factors contributing to this issue, namely, the markets in these environments are still evolving, companies do not see the economies of scale to establish units locally to supply the demand. Third, the technology is changing rapidly; it is practically impossible for educational institutions to keep pace with it for the reasons stated. And finally, there is a lack of awareness among these institutions as to what technologies exist outside their realm and how they can be acquired for the benefit of their students.
Despite years of extensive efforts in tackling the digital divide issue, organizations and initiatives have fallen short of delivering the benefits of IT to these developing environments. To assist the academic institutions in providing industry standard technologies, a framework for assessing the current solutions, identifying unresolved technical and business challenges and evaluating potential technological alternatives is needed. To develop this framework, the interdependencies among three key sub-systems academia, industry and government, in conjunction with the availability of low-cost technology alternatives must be explored.
Originality Requirement
The thesis describes a novel approach to develop a framework that applies learning from the MIT System Design and Management program to the analysis of the digital divide issue in academic settings in developing environments. An original examination of the existing initiatives addressing the issue is presented along with the exposition of major challenges that need to be further addressed to bridge the digital divide. A framework around information and communication technologies is presented along with the synergies needed among academia, industry and government.
In order to obtain tangible conclusions and provide concrete recommendations, the thesis applies the developed framework to India, which is a developing country, recognized as low-cost, high quality producer of information technology related products and services.
Content and Conclusions
This thesis approaches digital divide in the academic settings as a systemic issue in the developing environments. Prior work addressing the issue assumes that the problem can be tackled at an element level, through technology donations or by establishing state-of-the-art laboratories in a few select institutions, without understanding the principle dynamics at the sub-system and system levels. As a result low-cost alternatives which are optimized at an element level and which cannot be easily scaled have been developed. In general, helping academic institutions to keep pace with the clock-speed of information technologies and assisting the decision makers with the necessary understanding of interdependencies in the system are not addressed. The initial chapters of the thesis define the digital divide in academic settings in developing environments, explore and analyze the existing solutions and attempt to characterize the major challenges that are still unresolved. The middle chapters describe the methodology and approach taken to develop the framework and present a discussion on the systems approach to the problem. The final chapters apply the systems framework developed to an academic setting in India and discuss strategic implications and recommendations for overcoming the inherent tensions between quality and cost of education.
One of the primary conclusions from this research is that the supporting infrastructure and the policies surrounding indigenous software and hardware component manufacturing industries is a key ingredient in solving the digital divide problem. The recommendations address the roles of industry and government on how best to develop communication technologies and better equipped laboratories that strengthen the supportive infrastructure for academic institutions. In addition, important decision variables including the cost of education, initiatives for global technology awareness, use of low-cost technology alternatives, policies and direction for electronic component manufacturing industry and a common vision and partnership amongst academia, industry and government have been identified as the key building blocks for addressing the digital divide problem. Many of the issues in indigenous technology development can be overcome by employing system architecture principles to modularize and create products based on vertical product architectures which will facilitate easy adoption to the IT clock-speed. The proposed technology framework follows a modular architecture with open-standard programming interfaces. System engineering principles aid in defining the sub-system interfaces at various levels in a clear and terse manner. These principles are combined in a unique manner to develop a framework that simplifies the process of providing industry standard laboratories to students and researchers. This change in strategy calls for some inevitable cultural and systemic changes to the education delivery process and organizations in the academia.
Furthermore, this thesis demonstrates how existing approaches fall short of successfully addressing the technology gap issue due to their element level approach and presents a discussion on the implications thereof. In addition, it reiterates that equipping educational institutions with industry standard laboratories and training the students in the state-of-the-art methods are imperative to the developing environment’s future successes in the information technology industry.
System Design and Management Principles
The thesis draws upon a combination of SDM principles to achieve its objectives. Fundamental concepts from system architecture, system engineering, system dynamics, organizational processes, technology strategy and advanced software engineering are brought together to provide a comprehensive analysis of the issues surrounding existing approaches to solve the digital divide problem. In closely examining issues around the digital divide in academic institutions, important aspects such as weak component industry and manufacturing infrastructure, high-cost of technology imports, and lack of proper communication infrastructure are uncovered. The thesis examines a number of real-world implementations of industry-academia-government partnerships and draws lessons from these initiatives that have been incorporated into the proposed framework. A holistic systems perspective is required to understand the linkages between industry, academia and government and is adopted. Fundamental aspects of the academia-industry-government value chain are explored to identify these linkages and a framework that is an amalgamation of SDM principles, advanced technologies and key research findings is developed to aid decision makers in systematically approaching the digital divide problem.
In particular, knowledge gained from system dynamics and systems architecture courses, both of which advocate holistic thinking, was particularly valuable in these regards. The technology framework developed in the thesis extensively uses system architecture principles including identifying end users and their needs, defining product goals, analyzing the classic tension between cost and quality of education in designing a system that is modular, meets the low-cost requirement, provides industry standard software and hardware technologies to students through shared computational resource model and is extensible with small additional costs. System dynamics was employed to gain preliminary insight into the behavior patterns of sharing computational resources among partner institutions. Advanced software engineering principles are utilized to develop the design of the software system with object orientation, data abstraction and platform and protocol independence through the use of XML and other standards based technologies. Marketing strategy course provides valuable tools such as customer profiling, in this case, students’ computational and communication needs, to prepare taxonomy of technology in education. Financial and managerial accounting coursework assisted in developing a basic cost model for the proposed technology solution. A comparative cost-benefit analysis is presented that emphasizes the scalability, flexibility and affordability of the proposed technology framework.
Engineering and Management Content
By combining concepts from engineering and business, this thesis attempts to create a holistic view of challenges facing the academic institutions in developing environments with respect to acquiring and integrating industry standard software and hardware technologies into the curriculum. Engineering content includes technical review and analysis of initiatives addressing the digital divide issue and a comprehensive description of a proposed framework for deploying state-of-the-art mobile laboratories that leverage simplicity and affordability of wireless in local loop technologies. The author’s development experience in the field of telecommunications and wireless technologies provides unique insights into the challenges faced by indigenous software and hardware development industries. The proposed approach evaluates a fairly low-cost wireless enabled mobile laboratory, based on industry knowledge and technology trends, which would provide state-of-the-art hardware and software technologies to students and researchers in academic institutions in developing environments. Furthermore, it evaluates a resource-sharing model, based on grid computing, and describes a new grid service for enabling resource sharing among partner academic institutions.
Management content includes a characterization of the chasm that is created in the developing environments in adopting breakthrough technologies, an analysis of factors exogenous to the academic settings such as industry recruitment, training costs and cost of education and a note on the existing barriers to organizational change.
The thesis goes beyond design analysis “in the small” by taking a fresh look at how academic institutions need to change the way they approach the issue of providing industry standard software and hardware laboratories and recommends taking a big picture approach to gain better understanding of the issue and its implications. The creation of supporting infrastructure, policies and products for local markets is a daunting task that is rarely easy to accomplish but its benefits are enormous.
Statement of Authorship and Originality
The work performed to write this thesis is the author’s and is original.
Thesis Supervisor: Amar Gupta
Title: Co-director, PROductivity from Information Technology (PROFIT)
Dedicated to my parents...
Acknowledgments
I dedicate this thesis work in memory of my parents, Prof. S.B.Raghunathacharya and. S.B. Vijayalakshmi and my late grandfather Sri. S.B.L. Narasimhacharya. My father inspired me with his perseverance and integrity; my mother instilled in me aspirations to engender positive change and my grandfather showed me that knowledge is an ocean and learning should never stop. I am greatly indebted to them for all that they have given me. Today, I proudly carry their torch. If they were alive, I know that their joy would have had no bounds, and that they would have reflected on their son’s accomplishments with pride.
I thank the System Design and Management (SDM) program for believing in me and challenging my abilities at every level. I would like to extend my appreciation to Dennis Mahoney, Director of SDM program and Ted Hoppe for their help and kind words during times of distress. I have enjoyed the journey through this program and feel fortunate to have made friends with some of the brightest minds among my classmates.
I would like to sincerely thank my advisor Dr. Amar Gupta who constantly challenged me to raise the bar, guided me, kept my focus and was always available to exchange ideas that helped formulate this thesis. I am grateful to my manager at Nortel Networks, Franco Travostino and my project leader Inder Monga for their help throughout the program.
I like to also thank our friends Prashanth, Nagashree, Anjali, Sridhar, Srikanth, Vani, Srinadh, Padmaja and Geetika, for their support and for still continuing to be friends with us!
Above all, I offer my love and gratitude to my wife Kiran and daughter Raaga Manjusha, I would not have dared to take on this wonderful adventure if not for their immense love and sacrifice.
Contents
List of Figures 11
List of Tables 12
Chapter I: Introduction and Overview 13
1.1 Introduction 13
1.2 Academic Settings in Developing Environments 17
1.2.1 Academic Culture and Change Management 18
1.3 Thesis Outline 19
1.3.1 Tactical and Strategic Objectives 19
1.3.2 Project Considerations 20
1.3.3 Thesis Organization 20
Chapter II: Project Objectives and Background 22
2.1 Providing Industry-standard Technologies to Students in Developing Environments 22
2.1.1 Problem Statement 22
2.1.2 Tactical Objectives 23
2.1.3 Strategic Objectives 24
2.2 Academic Environment 24
2.3 Drivers for Growth in Educational Technology 25
2.4 Bringing Industry-standard Laboratories Inside 25
2.5 Metrics 26
2.5.1 Technology Spending in Education 26
2.5.2 Cost per Student 26
2.5.3 Targeted Education Spending 27
2.6 Problem Scope 27
Chapter III: Methodology and Approach 28
3.1 Overview of the Proposed Framework 28
3.2 Discovery Process 30
3.2.1 Market Research 30
3.2.2 Stakeholder Needs Analysis 31
3.2.3 Taxonomy of Technology in Education 31
3.3 Benchmarking 33
3.3.1 Simputer 33
3.3.2 Thin-client Thick-server Paradigm 34
3.3.3 Infothela 35
3.3.4 Digital Gangetic Plain 35
3.4 Relevant Literature Search 35
3.4.1 High clock-speed 35
3.4.2 Technology TCO 36
3.4.3 Technology Unavailability 37
3.4.4 Impediments to Innovation Adoption 37
3.5 Populating the Academia-Industry Partnership Model 38
3.5.1 Student Enrollment 38
3.5.2 Research Budget 39
3.5.3 Institution’s Innovation and Intellectual Property 39
3.5.4 Institutions’ Reputation 39
3.5.5 Research Grants 40
3.5.6 Academia-Industry-Government Value Chain 40
3.5.7 Recruiting 41
Chapter IV: Systems Approach to the Solution 42
4.1 Real-world Implementation of System Engineered Solutions 42
4.1.1 Leveraging Computer and Television Assets in Brazil to Deliver Educational Content 42
4.1.2 Media Lab Asia 43
4.2 Systems Thinking 45
4.3 Proposed Technology Framework 46
4.3.1 Internet Technologies 48
4.3.2 Wireless Technologies 48
4.3.3 Grid Computing 49
4.3.4 Mobile Wireless Enabled Laboratories 51
4.3.5 Architecture Design for a Grid Service Running in Mobile Laboratory 56
4.3.6 Point-to-Point Shared Online Laboratories 64
4.4.1 System Dynamics Model 66
4.4.2 Sensitivity Analysis 68
4.4.3 Simulation Results and Discussion 72
4.4.4 Emergent Dysfunctional Properties 72
4.5 Proposed Business Framework 72
4.6 Bringing the Framework Together 73
Chapter V: Applying the Framework to India 74
5.1 Overview: Three Environments 75
5.2 Technology in Education 77
5.3 IT Outsourcing in the Perspective 81
5.4 Strategic Direction for Indian Technology Industry 82
5.5 Application of the Framework 83
5.5.1 Education and Research Network, India 83
5.5.2 Use case scenario for cost comparison 84
5.5.3 Wireless Enabled Mobile Laboratory Service 86
Chapter VI: Strategic Recommendations and Conclusions 88
6.1 Grid Computing 88
6.2 Wireless Last-Mile Connectivity 88
6.3 Understanding the Nature of the Technology Gap 89
6.4 Sources of Digital Divide 89
6.5 Readdressing Roles of Academia and Industry 89
6.5.1 Rule of 10s and Concurrent Training 90
6.6 Reevaluating Technology Gap in Developing Environments 90
6.6.1 Industry in the value chain 90
6.6.2 Process Improvement 91
6.6.3 Enterprise Approach to Technology Gap 91
6.6.4 Metrics 91
6.7 Academic Institutions in IT Centric Economies 92
6.8 Conclusions 92
Appendix A: Terminology 95
Appendix B 98
Bibliography 100
List of Figures
Figure 1-1: Technology Diffusion in Educational Institutions 17
Figure 3-1: Clock-speed of the Electronic Chip Industry 36
Figure 3-2: Academia-Industry-Government Value Chain 41
Figure 4-1: Upstream and Downstream Influences in Architecting a System 45
Figure 4-2: Circles of Complementary Enabling Technologies 48
Figure 4-3: High-level Grid Architecture and Functional Blocks 50
Figure 4-4: Prototype Architecture for a Wireless Enabled Mobile Laboratory 54
Figure 4-5: State-of-the-art Mobile Laboratory 56
Figure 4-6: Lab Service System Model using Grid Infrastructure 57
Figure 4-7: XLab Grid Service Policy Interactions 60
Figure 4-8: Protocol for the XLab Service Application to use Grid Services 63
Figure 4-9: MIT’s Shared Online Laboratories, WebLab 4.0 Architecture 65
Figure 4-10: Reference Modes for the Proposed System Framework 67
Figure 4-11: Technology Sharing Among Participating Institutions Model 68
Figure 4-12: Perceived Quality of Education 69
Figure 4-13: Actual Technology Sharing 69
Figure 4-14: Cost of Education with respect to technology spending 70
Figure 4-15: Effect of perceived quality on technology upgrade pressure 71
Figure 4-16: Business Aspects of the Proposed Systems Framework 73
Figure 5-1: Indian Education System with Thesis Focus Segments Highlighted 76
Figure 5-2: PC Penetration into Educational Institutions in India 77
Figure 5-3: Internet Growth in India 78
Figure 5-4: The Cost of Network Infrastructure in India 79
Figure 5-5: Evolution of Offshore IT Outsource Model and its Impact on Value Creation 82
Figure 5-6: Education and Research Network, India 84
List of Tables
Table 3-1: Taxonomy for Different Technologies in Education Based on Their Needs 33
Table 3-2: Technology Infrastructure in a Developing Environment 38
Table 5-1: PC Penetration into Educational Institutions 78
Table 5-2: Internet Growth in a Developing Environment 78
Table 5-3: Approximate Cost of Technology Infrastructure Installation 85
Table 5-4: Costs of Deploying a Wireless Enabled Mobile Laboratory 86
Table A-1: Terminology 97
Table B-1: Indian Software Industry Growth (Source: NASSCOM) 98
Table B-2: Shared Technologies System Dynamics Model Equations 98
Chapter I: Introduction and Overview
This section describes a classic digital divide issue between the technology haves and the have-nots in the academic setting. The issue concerns the inability of academic institutions in developing environments to keep up with the high acquisition and ownership costs, and rapidly changing industry standard software and hardware technologies to create the best-in-class learning experience for their students and researchers. This thesis, attempts to provide a framework to demonstrate how academic institutions in developing environments could leverage emerging technologies to mitigate the digital divide; in particular the notion of grid computing and wireless connectivity in the last mile are explored as unifying force between developing and developed environments. The framework developed encompasses a discussion around the enabling technologies. Furthermore, it recognizes the importance of a systems perspective of the digital divide issue; it addresses the organizational and business issues surrounding the digital divide problem.
1.1 Introduction
Digital divide is an age-old technology and management topic. Nonetheless, it remains elusive to organizations that attempt to mitigate it. Digital divide issue in education can be approached in many different fronts. Some of the previous approaches taken are in the areas of providing technology and infrastructure donations, establishing funds and organizations to aid technology adoption, promoting local inexpensive technology alternatives, conducting educational seminars, games and conferences and enabling strong interconnections among local and global centers of excellence. The author believes that approaching the problem in any one front will not solve the problem at hand. Taking a system of systems perspective is important to successfully address an issue faced by over two-thirds of the world’s academic population. Specific initiatives, their pros and cons, are listed in Table 1-1.
|Category |Initiative |Pros/Comments |Cons/Recommendations |
|Technology and |Reuters Foundation, |Initiative to bring together academic, |The project rightly recognizes |
|Infrastructure |Stanford University [35] |corporate and NGO sectors. It gives |Governmental assistance in setting |
|Donations | |technology experts an opportunity to |the right policies; but, it doesn’t |
| | |come to Stanford campus and apply their|involve any government |
| | |vision and talent to address the |representatives or enforce policy |
| | |challenges faced by developing |change. |
| | |environments. | |
| | |Good set of corporate sponsors |Little local industry involvement. |
| | |including Microsoft, Cisco, HP, Philips| |
| | |and others. | |
| | |Moderate number of projects ranging |None of the current projects is |
| | |from strengthening financial |looking into improving education or |
| | |infrastructure and e-commerce in India |enabling technologies. |
| | |to Satellite imagery and GIS data for | |
| | |agriculture. | |
| |Computer Aid International|The world’s largest non-profit supplier|Computers donated are older versions.|
| |[40] |of computers to developing |No application software support. |
| | |environments. Support from the | |
| | |government and industry. | |
| | |Focused on the high-schools. |Not directly affecting the |
| | | |engineering education arena. |
| | |Training in computer repair and usage | |
| | |to people from developing environments.| |
|Funds and organizations|Media Lab [41] |Programmable Lego bricks and MPEG video|One of the three planned divisions |
| | |encoding standard are two successful |which was supposed to be education |
| | |endeavors in 25 year history of Media |(edevelopment) is still-born. It has |
| | |Lab. |not materialized. |
| | |Support from MIT and local governments |Little impact on the education and |
| | |in terms of funds, domain experts, |technology infrastructure in rural |
| | |research, students and faculty. |parts of Asia. The government is |
| | | |rethinking their strategy around |
| | | |Media Lab Asia [67]. |
| | [26] |Concept is to provide a web medium that|To date, no significant impact on the|
| | |will bring together visionaries from |targeted communities. No focus, which|
| | |around the world to design solutions |is good for creativity but very |
| | |for real-world problems in developing |detrimental to achieve anything |
| | |environments. Good concept. |concrete. Works in the education |
| | | |field are very premature with little |
| | | |or no impact. |
| |IITK BRiCS [42] |Introducing state-of-the-art technology|Dependent on importing Lego kits |
| | |and concepts into high-schools and |which are very expensive for an |
| | |colleges through games and robotic |average educational institution. |
| | |competitions. | |
| | | |Must find ways to build alternate |
| | | |low-cost robot kits. For example, |
| | | |using wires for gears has proved to |
| | | |be viable. Similarly replacing the |
| | | |Lego programmable brick with low-cost|
| | | |Handyboard [68] may help. |
|Collaboration between |American Indian Federation|Innovative partnership model between |Not scalable unless critical momentum|
|Global Centers of |[43] |American and Indian organizations |is gained in establishing similar |
|Excellence | |striving for similar goals. Established|ties. No technology orientation; does|
| | |school for 400 child laborers. |not have any impact on the |
| | | |engineering education. |
|Indigenous Innovation |TeNeT Group [3] |Excellent model that will dramatically |Working toward government support, |
| | |impact the way educational institutions|must build coalitions with like |
| | |acquire technology and the costs of |minded organizations in India as well|
| | |ownership. |as abroad. |
| | |Proven technology, deployed | |
| | |successfully in India and some African | |
| | |environments. | |
| | |Right approach; it involves the local |In the near term, building |
| | |industries, local talent, local |relationship with low cost |
| | |government and has a global reach. |manufacturers in Taiwan and China may|
| | | |help. In the long term, drive local |
| | | |manufacturing infrastructure, |
| | | |policies and applications. |
Table 1-1: Digital Divide Initiatives and their Pros and Cons
As developing environments approach the problem of digital divide more systematically, they will find the tension between balancing their educational infrastructure through international aid and assistance with low-cost, highly available local IT providers. The investment in technology infrastructure for the academic sector is rarely justified by tangible returns.
Institutions have approached this battle by instilling an innovative culture. The innovative culture framework includes generating hi-tech solutions to real-world problems and creating value to the industry, government and the community. This approach has generated millions of dollars of research grants to select institutions from various sources, including the government and the industry. It propelled many governments in developing environments to champion exodus from old-way of doing things into making hi-tech the center of their universe [44]. Despite the efforts, many recognize the need for a cohesive framework [31] that addresses how to scale the model to academic institutions in the developing worlds. In short, how can the academic institutions in developing environments generate the level of talent required for further strengthening the country’s foray into Information Technology?
The following sections give an overview of the educational sector and the culture in developing environments and an overview of thesis, which will provide a framework to answer the question raised above. Although significant detail on several developing environments is provided, the thesis focuses on Indian academic setting because the framework developed will be applied specifically to India to gain concrete insights and conclusions.
1.2 Academic Settings in Developing Environments
Typically the technology advances have been adopted by the most developed environments where some of the recent innovations such as the Internet, Wireless and Grid computing originated.
[pic]
Figure 1-1: Technology Diffusion in Educational Institutions
The developing environments tend to be part of the delayed adopters (Chasm B) as shown in the “crossing the chasm” concept [19] in Figure 1-1. Therefore, these environments have very little impact on how the technology should look (form) or behave (function) to suit the local cultural, technological and business needs. What would be beneficial to the developing environments is if they are able to take part in the development of a new technology as early or lead adopters. This thesis attempts to take one step in that direction by proposing to introduce wireless connectivity and Grid computing to the students and researchers in developing environments. In the next section, a brief overview of the academic settings in three developing environments is presented.
1.2.1 Academic Culture and Change Management
The cultural aspects of academia cannot be overlooked if one aspires to bring change into the way these institutions think and perform. Organizations that have established offices or developed initiatives in Indian schools recognize the importance of understanding the local academic culture and accordingly place emphasis on the aspect. For example, in India the academic environment is very competitive and the method of instruction is mostly lecture based versus an interactive environment for open dialog and exchange of ideas. The organization structures are hierarchical and typically seniority leads to career advancement for the faculty. Research is not on the primary agenda for a typical academic institution; thereby interaction with the industry and the government practically does not exist. The lack of interdependence among the government, industry and academia, except for the purposes of obtaining funds, creates a possible chasm between these organizations in terms of their understanding of each others culture, vision and synergy.
The cultural shift must occur in several fronts; first one needs to understand the value of technology in education; second adapt to the changes in the industry, third, understand the past, set a common vision and direction; fourth create a sense of urgency; fifth support a strong leadership role; sixth align the political sponsorship and develop the implementation plan; and finally create the enabling structures and communicate the change to involved stakeholders [12]. This thesis takes the aforementioned gated process steps and proposes a change management strategy in the academic settings to deploy the developed framework.
1.3 Thesis Outline
This thesis provides an analysis of the digital-divide issue specific to the academic arena. By addressing the fundamental aspects such as information and communication infrastructure, better equipped laboratories and industry and government support to academia, a low-cost mobile laboratories model is developed. The model provides a framework for discussing the tactical implications of low-cost solutions to provide industry standard laboratories to students and the role state-of-the-art technologies in education must play in the future successes of India, its industries and academia.
1.3.1 Tactical and Strategic Objectives
Historically, the majority of educational institutions in developing environments provide software and hardware technologies that have been acquired very early on when the laboratories were established. In many of these institutions there is lack of systematic approach to acquiring newer technologies to keep the students abreast of advancements [50]. Reasons for not updating to newer state-of-the-art hardware and software technologies are many as noted in the problem statement.
The thesis attempts to reevaluate the issues and present a framework based on newer communication technologies and resource sharing alternatives being developed by the Grid computing community. The Grid computing approach is similar to the fundamental idea behind an academic institution, which is to share multiple resources in a cohesive fashion in order to impart knowledge to large numbers of students simultaneously.
In specific, the thesis will develop architecture for a wireless communication infrastructure to fulfill the last-mile problem that is faced by majority of the educational institutions in developing environments. In addition to the proposed communication infrastructure, access to industry standard software and hardware technologies must be provided. Grid computing infrastructure being developed by the Global Grid Forum [71] attempts to establish standards around resource sharing, protocols, security and policies which will enable organizations around the globe to share their computational resources with other organizations as if the participating entities are all part of a bigger virtual organization.
This thesis attempts to leverage the Grid infrastructure [72] and proposes a new laboratory grid service called XLab, an extensible laboratory service, and defines the architecture, functional and design specifications including functional module description, internal and external interfaces, and application programming interfaces (API) needed to implement the service. Furthermore, a cost model of the proposed solution is developed and a comparison with a traditional solution is presented. At the organization level, the proposed strategy to combine wireless technologies and grid computing best leverages the resources among partner institutions to provide low-cost, state-of-the-art laboratories to their students. Enterprises benefit because they will have an opportunity for hiring graduates who are proficient in industry standard software and hardware technologies, which translate into shorter learning curves, higher productivity and faster product development cycle time.
1.3.2 Project Considerations
This thesis not only leverages the classic literature and earlier work in this field but also makes an effort to recognize the organizational dynamics and strategic and tactical implications of the framework. Leadership, academic culture, industry trends, system dynamics and technology strategy add depth to the discussion and the resulting framework.
1.3.3 Thesis Organization
First, the thesis takes a closer perspective on the problem, affecting factors and important metrics, in order to understand the project background and formulate the objectives. In the methodology and approach section, the approach taken to tackle the issues in a novel way as well as the methodology used to leverage prior art for coming up with a framework are discussed. Once the framework has been defined, the section on systems thinking presents the possible dynamics in a system where the shared resources model in academia is analyzed.
In Chapter 5, the framework is applied to the Indian academic environment. Recommendations and strategic results section discusses how the framework attempts to bring industry standard technologies to students in the developing world and its implications on the students, institutions and the industry. The conclusions section highlights lessons learned from the context of academic institutions in India and key findings. There are a number of acronyms and technical terms used throughout the document which are documented in Appendix A; related material is documented in Appendix B.
Chapter II: Project Objectives and Background
An SDM thesis must satisfy many requirements, one of the most important one being a balanced approach between engineering and management aspects and addressing them with a clear systems thinking in perspective. The thesis framework and objectives recognize the importance of systems thinking and ensure that the issues are approached bearing a holistic view in perspective.
2.1 Providing Industry-standard Technologies to Students in Developing Environments
A number of global organizations have taken measures to provide better educational facilities to kinder garden, primary and middle school level students. For example, the World Bank [14] initiated District Primary Education Program (DPEP) and Higher Education Projects [29] in India in the 1990s to help provide children ages 6 to 14 to get quality primary education. So far it has reached 60 million children and cost US $1.2 billion. Non-profits like Jiva [24], Digital Dividend [15] have taken a different approach for providing state-of-the-art education skills through the establishment of local community centers. Media Lab Asia [39] aspires to help provide cheap, reliable, innovative computing to the rural areas of environments such as India. However, very little has been done to provide adequate Internet facilities and reliable experimental tools to students of engineering education. The main objective of the thesis is to enable engineering colleges with communication and computational infrastructure at an affordable cost.
2.1.1 Problem Statement
The author believes that students learn best when they practice the concepts learnt in the classroom through experiments and hands-on exercises. However, in many developing environments and some rural areas of developed environments, the academic institutions are simply unable to provide industry standard laboratories to their students. There are several factors contributing to this inability. First, given the strength of the local economies and the buying power of the currencies, acquiring these technologies is extremely expensive. Second, many of these improved technologies are not easily available. There are several factors contributing to this issue, namely, the markets in these environments are still evolving and the manufacturing companies do not see the economies of scale to establish units locally to supply the demand. Further, the technology is changing rapidly; it is practically impossible for educational institutions to keep pace with it for the reasons stated. And finally, there is a lack of awareness among these institutions as to what technologies exist outside their realm and how they can be acquired for the benefit of their students.
While many potential projects came up during the initial discussions with students and faculty in developing environments and with the advisor alike, it was quickly evident that the problem lies at a much deeper level. Although educational institutions in India recognize the issue of poor laboratory facilities, they are unable to leapfrog to the next best alternative due to insufficient understanding of the implications of the problem at hand and lack of the holistic systems view. By understanding the key drivers for the need to enhance laboratories facilities and for improved communication infrastructure, efforts can then be best deployed. Using the framework of low cost laboratories model, the following tactical and strategic objectives are identified.
2.1.2 Tactical Objectives
The problem of digital-divide in general is well understood by many of the developing environments. For example, in India some states are adapting to e-governance and other initiatives to bring the benefits of technology to the masses in rural areas. Traditionally, the digital-divide in the academic setting has been viewed with little attention. It is now becoming more evident how critical it is to have students exposed to strong engineering and manufacturing skills for the future success of the country. The following barriers to creating industry standard technologies locally have been identified [5].
• Weak ties between the research wings and the industry
• Very disorganized and poor electronic component industry
• Non-aligned incentives for indigenous manufacturers to work with local companies
• Weak policy around the hardware design and manufacturing industry
All of the above factors affect the local educational institutions’ ability to acquire industry standard technologies. A closer look at the technologies deployed in most of the institutions today for internet connectivity reveals that they are not only expensive and inefficient but they are not scalable too.
2.1.3 Strategic Objectives
Once the tactical objectives have been addressed, it becomes important to understand what it takes to implement a system based on the framework in a real-world. There are various aspects to be given serious consideration.
Who will implement the service, to whom will the services be provided and who will pay for the services?
How should the educational institution approach the total cost of ownership issue?
How indigenous companies are encouraged and their products are leveraged?
What are the policies that will govern the successful working of the proposed system?
The following sections detail other exogenous variables that affect the thesis objectives and influence the results of the framework including the academic environment, drivers for growth of technology in education, industry partnership and key metrics.
2.2 Academic Environment
Indian education has traditionally been weak in imparting hands-on engineering education while extremely strong on other aspects. The government has taken steps in the recent years to strengthen the hardware industry; however, there seems to be a lack of consistency and commitment in the government policy [52]. There is a strong need for improving the engineering and manufacturing [51] facilities to jump-start local market for indigenous software and hardware products.
2.3 Drivers for Growth in Educational Technology
Technology has become an inherent part of the educational systems in the developed environments. The use of computers and communication technologies in the classroom and on campus is taken for granted in majority of these environments. However, developing environments are facing a challenge in providing software and hardware technologies to their local academic institutions. The application of these technologies in daily coursework, classroom or on campus is still a distant future for these institutions for the reasons noted in the earlier sections.
There is a steadily growing demand for graduates with expertise in hi-tech methods and tools. In the US, 6 out every 10 people have computers and 1 out of every 2 persons uses the Internet [69]. The situation is different in developing environments; for example, in India 1 in 1000 students has access to computers and 7 in 1000 people use Internet [25]. There is a great upside potential for technology growth in the academic settings in India given the above statistics. Besides, the effects of globalization, increasing investments by multinational corporations and affordable alternate technologies such as wireless, all help boost the need for a strong technology infrastructure in educational institutions within developing environments.
2.4 Bringing Industry-standard Laboratories Inside
We noted how critical it is for developing environments to invest in educational technologies in the classroom to better serve the demand for technology oriented employment opportunities being generated and the need for building skills in the engineering and manufacturing of complex systems. Several of the top engineering colleges in India have established strong ties with industry leaders such as Intel and HP which setup sophisticated microprocessor and other electronic component fabrication laboratories on campus. Needless to say, this model needs to be scaled further to build out a human capital base of qualified electronic fabrication, design and manufacturing expertise. Industry needs to partner and work closely with the academia to provide the right infrastructure and environment for the students to gain knowledge and expertise in the desired fields.
2.5 Metrics
Several important metrics must be considered to make the case for using industry standard technologies in an academic institution, and the benefits to its students, to the institution and to the industry. This section describes three important metrics that need to be carefully studied to clearly understand the benefits these educational technologies bring, the costs involved, and how and where to target the limited resources that an institution has at its disposal.
2.5.1 Technology Spending in Education
Technology spending in many educational institutions in developing environments is a small percentage of the overall education budget. For example, when compared to a mid-tier academic institution in the US, a peer institution in the developing environments will stand a distant chance of competing on the basis of technology infrastructure. Over the years, the costs of technology ownership have come down significantly and many schools in the US have benefited directly as a result of this trend, however, these same trends have not translated into any significant benefits to institutes of higher learning in developing environments. There is a need for taking a fresh approach to this issue, software and hardware industries in these environments have to be strengthened and encouraged to innovate for the local markets.
2.5.2 Cost per Student
The Indian educational system employs a low cost structure which makes it affordable for students; Cost of education has been rising steadily, in the early 1990s the average tuition and fee was roughly US $50-$75 per student per year while in the US the average cost per student per year during the same time was US $15,000 [72].
The supporting infrastructure for student loans, scholarships, assistantships and other fellowships helps the students in the US cover their educational expenses. In the developing environments, the supporting infrastructure is weak. Since the education costs are not very high, the banks and other lenders are not attracted to invest in such programs and credit tracking and rating systems are not in place as well. The governments are unable to help the students as much because of lack of funds in addition to the prior stated reasons.
2.5.3 Targeted Education Spending
Academic institutions need to be targeted towards where and how they spend their limited resources. In many cases it may be appropriate to expend resources in acquiring tools, technologies and skill-set in the high-growth industries where there will be a substantial return on their investment in terms of student satisfaction, employment opportunities for their students, obtaining research grants and so on.
2.6 Problem Scope
What is lacking in many of the alternative solutions addressing the technology gap in developing environments is that, the solutions do not address the fact that technology is changing rapidly and the costs of ownership of new technologies to the academic institutions is extremely straining on their budgets. This thesis research focuses on bringing the state-of-the-art software and hardware systems to students through the use of three complementary technologies.
1. Internet related technologies and Virtual Private Networking
2. Wireless Networks and,
3. Grid Computing
A cost model is developed around the proposed design and is compared with an alternate existing solution.
Chapter III: Methodology and Approach
The methodology and approach to address the tactical and strategic goals can be categorized broadly into six phases. Phase I involves the discovery process of technologies in use today in developing environments’ educational systems which includes market research and identifying stakeholder needs and developing a taxonomy of the communication and information infrastructure. Phase II develops an understanding of the current models and their approach to the issue and perform relevant literature search and analysis. Phase III populates the Industry-Academia partnership model by taking a deeper look into some key variables such as student enrollment, student satisfaction, research budget and so on. Phase IV takes a systems approach to devise a solution framework, Phase V applies the framework developed to Indian academic setting and the final phase discusses the strategic and tactical implications and conclusions from the framework.
3.1 Overview of the Proposed Framework
In systems composed of many interacting feedback loops and long time delays, the cause of an observed symptom may come from an entirely different part of the system and lie far back in time [22]. It is important to understand the system in which educational institutions operate and the different sets of feedback loops to understand the issues at hand. The framework developed in this thesis takes a systems approach to the problem. It looks at Industry, Academia and Government as three sub-systems and how their interdependent roles affect the educational system. The thesis takes a look at the existing synergies among the three sub-systems from a developing country perspective and evaluates a new model in which the sub-systems behave as interdependent entities that recognize the feedback loops. Some environments have demonstrated how the three sub-systems must operate to successfully leverage the synergies. For example, in Japan, government, industry and academia have collaborated on several projects in the past [83]. Together they are working to increase productivity, foster the growth of new business and lay a firm foundation for the resumption of healthy economic growth [82] through a collaborative effort that includes some of the large private sector members including Toshiba, NTT DoCoMo, Canon and University of Tokyo. A few of the key recommendations from the commission are to encourage government to reduce the depreciation periods for high-tech equipment, increase tax credits for R&D investments, increase spending on basic research activities, set minimum level of regulations necessary for competitors to enter and establish their businesses and promote innovation and entrepreneurship.
On the similar lines, by 1997 the academia-industry partnerships had resulted in 2362 joint research programs up from a meager 56 projects in 1983; the first two big segments among these projects are hardware (307) and software (221). There are several problems with the joint research programs. First, all the funds have traditionally come from the private sector, but this is changing to a model where the academic institution matches the funds. Second, intellectual property rights belong to the academic institution; efforts are being made to give priorities to the partner company. Third, the procedures for contract renewal are too complicated, simpler methods must be adopted. Commissioned research programs increased from 1286 in 1983 to 4499 in 1997, while the research grants to academic institutions grew from 2.2 billion yen to 33.3 billion yen during this period. A striking observation reveals that more than 150 billion yen in research grants were made by the Japanese private sector to Universities abroad; this amount is more than double that granted to Japanese Universities. The number of patents registered in Japan for 2001 totaled 12580, among which only 161 belong to Japanese Universities [84].
The proposed Industry-Academia partnership model traces the effects of several important attributes of the system on the effectiveness of the solution. Educational institutions feed valuable research outcomes and insights into the Industry, thereby strengthening the growth of the industry in the right direction. The industry benefits from bringing in exceptionally qualified students to propel its growth further through internships and campus recruiting. There is a strong interdependency between these two entities in the value chain that must be nourished well in order to increase student satisfaction, student enrollment, research and innovation in the educational institution as well as breakthrough technologies, leading products and growth for the industry partner. Furthermore, the industry in return benefits by reducing the costs of initial training and bringing the student recruits up to speed on their technologies. The first three phases are discussed in this section.
Previous work addressing the technology gap in educational institutions in developing environments involved lot of words, but resulted in meager implementation. Recent efforts [73] by Jhunjhunwala [3] and Dhande [53] are notably different from the previous efforts because they are a step in the direction that Japan has taken as detailed earlier. These efforts are not only helping their students participate in advanced engineering design but are already creating value to people in developing environments by creating their ICT infrastructure at an affordable cost. There are numerous initiatives in the direction of helping the rural communities benefit from the technology revolution [24]; however, there are language and cultural barriers that need to be simultaneously addressed as well. Prior study however brings one key insight into the proposed framework. For successful proliferation of technologies into the developing environments, developing low-cost technologies and considering regional culture and language aspects are both extremely important.
3.2 Discovery Process
The discovery process was divided into understanding the market, issues and existing solutions, studying the stakeholder needs and preparing taxonomy of technology in education.
3.2.1 Market Research
Use of technology in the educational institutions has become an interwoven aspect of the curriculum. The question is no longer if technology helps provide better learning but it is one of how well one can leverage technology to impart world-class education as well as foster cutting-edge research activities.
It should be noted that the PC and Internet related technologies used in the educational institutions in some of the developing environments are as old as a generation while some of them indeed have the latest and the best available technologies; the latter is an exception and not a norm.
It is argued by some experts as the inevitable “Turning the bottom of the pyramid upside down” [91, 7, 8, 15] and very attractive market to go after, and for various other reasons such as lower cost of labor, access to centers of excellence and so on. More specific market research details are provided in Chapter 5 in the context of India.
3.2.2 Stakeholder Needs Analysis
Students, researchers, educators, educational institutions, administrators, industry partners, local governments and multinational corporations are some of the key stakeholders. Although there are several different needs that each stakeholder may have, there are a few needs that span the entire spectrum and are common to all of them, important ones among them are, enhanced quality of education through hands-on training, access to technologies that will not only enable innovation and creative problem solving but also add to the intangible assets such as reputation of the stakeholder and perceived value of the institution among the local and international communities.
3.2.3 Taxonomy of Technology in Education
The focus of this section is on developing taxonomy for engineering education. Each individual educational institution has multiple needs in using technology in education. The requirements vary among the institutions based on the kind of training and research work being conducted at that institution. The requirements also vary among different kinds of applications that the students in the institutions use. This section captures the essence of these requirements and categorizes the need for ICT and the extent of deployment.
|Category |Technology Requirements |Specific Aspects of Education |Enabling Technologies |
|High-end research |High-end computational resources |To enable researchers in the |Supercomputers from CDAC Servers |
|organizations and |including supercomputers, powerful |engineering field to perform |and workstations from local |
|institutions |servers, workstations, robust and |advanced research including |computer vendors, high-speed |
| |highly reliable power systems, access|bio-engineering, distributed |optical backbone networks based on|
| |to information on state-of-the-art |and parallel processing, image |disruptive technologies from TeNeT|
| |research and high-bandwidth internet |processing for weather |group and other local equipment |
| |connectivity |prediction, artificial |manufacturers and carriers such |
| | |intelligence. |BSNL and MTNL. |
| | | | |
| | |Software designed for the above| |
| | |mentioned specific applications| |
| | |such as AccuWeather, MIT’s | |
| | |exokernel etc. | |
| |Sophisticated software tools for | |Open source Linux, Solaris for |
| |number crunching and analysis, | |education, ’s office|
| |multitasking operating system | |productivity tools, other powerful|
| |environments, regression tools, | |software tools either developed |
| |database tools, artificial | |in-house or acquired from FSF and |
| |intelligence tools and reporting and | |. |
| |presentation tools. | | |
|Mid-tier technology |Excellent Information and |First, provide adequate |PCs from local computer vendors, |
|education institutions|Communications infrastructure to |communication infrastructure to|last-mile connectivity through |
| |support effective pedagogy. |engineering students so that |various up and coming technologies|
| |Computational resources to provide |they can now perform research |such as DSL, Cable Modem, Wi-Fi, |
| |enhanced hands-on learning, access to|and collaborate with peers. |VSAT and WiLL. |
| |world-class educational and research |Second, provide operating | |
| |resources including faculty and |environment to allow computer | |
| |students in other educational |science and electronics | |
| |institutions. |engineering students to develop| |
| | |programming skills. | |
| |Software tools to provide added value| |Open source Linux, Solaris for |
| |to in-class training. Software | |education, ’s office|
| |operating environment that provides | |productivity tools, other powerful|
| |exposure to various kinds of | |software tools either developed |
| |real-world applications and work | |in-house or acquired from FSF and |
| |settings. | |. |
|Primary technical |A good collection of low-to-medium |This segment is not currently |Local vendors, used equipment and |
|skills development |capacity computational resources that|the focus of this thesis |software from organizations such |
|institutes |enable simple yet powerful |although it is stated here for |as computer aid international, All|
| |introduction to technology and its |presenting the complete |India Federation etc. |
| |applications. |picture. | |
| |
|Table 3-1: Taxonomy for Different Technologies in Education Based on Their Needs |
3.3 Benchmarking
Benchmarks of four other technologies give a view of how others are trying to address the same problem. To understand the magnitude of the problem better, let us first take a look at how others have approached the digital divide problem.
3.3.1 Simputer
The Simputer [54] is a low cost portable device alternative to PCs, by which the benefits of IT can reach the common man. It has a special role in the developing world because it ensures that the illiteracy barrier to handling a computer is eliminated through the voice-enabled interface. However, its applications to people who are uneducated are not very clear at this point in time, given that there is little content on the Internet that is designed with this category of audience in perspective. Localized content that is usefully to such an audience must be developed and propagated for this technology to have the intended impact.
Simputer concept has its roots in the belief that bridging the digital divide can be achieved through simple, shared interfaces based on voice, sight and touch. The focus of this solution is more on the rural parts of the country, although the technology could be of tremendous value to students in terms of gaining access to Internet for file sharing, email communication, researching etc.
3.3.2 Thin-client Thick-server Paradigm
A few organizations are taking a look at the client server technologies of the 1990s and how they can help solve the TCO problem for educational institutions. Emergic Freedom [55], based in India, is one such company driving this initiative of Thin-client and Thick-server solution. There’s nothing new about this solution. It was the precursor to the client-server technologies that has come to be used in almost all the software solutions over the past decade. The issue with this is that there is need for excellent network infrastructure for robust scalable implementation. It would work well in a Local Area Network scenario but would depend heavily on the network infrastructure in a Wide Area Network scenario.
3.3.3 Infothela
Infothela [34] is a mobile integrated platform supporting voice, video and data transmissions without the use of electricity and wired network infrastructure. This is very well suited to the rural parts of India where telecom and power suppliers do not have any incentives to enter and lay the infrastructure. The focus again is on bringing computing, Internet, and telephone facilities to the rural communities.
3.3.4 Digital Gangetic Plain
This project [53] is aimed at extending the widely know Wi-Fi (802.11, 802.16) network beyond its traditional reach of few hundred feet into a reach of over 40 Km. This will be able to provide Internet connectivity to the remote places where the land line infrastructure may never materialize due to the economies of scale.
3.4 Relevant Literature Search
The existing work in this area touches upon four key aspects of the framework, high clock-speed of the technology industry, total cost of ownership in technology for educational institutions, technology availability and impediments to innovation. The following sections take a look at each of these aspects.
3.4.1 High clock-speed
Most technology companies rely on speed as a major differentiator. There’s speed in the introduction of newer technologies to market, there is speed in the product development process as well as in the internal management’s decision-making process. Fine [9] coins the term clock-speed to address the speed at which companies and industries change in terms of product, process, and organization. The Figure 3-1 shows high-clock speed of the electronic chip industry. We have seen how this plays into most of the electronics industry, right from the PC to the digital camcorder. On one hand, it has changed the way people learn, interact and communicate; on the other hand it has increased the digital divide among between developing and the developed nations’ educational institutions. Speed may be good for the Industry for several reasons including for the race to differentiate from competition, but it sometimes has a high toll on its consumers. Just as many consumers are hesitant or incapable of upgrading to the newer technologies due to high costs of migration in terms of learning curve, additional training and technical issues such as downtime etc., most educational institutions find it hard to upgrade their technology infrastructure to the industry standard.
Some may argue that it is not necessary for academia to possess the latest and the best available technology. However, the argument does not take into consideration that providing state-of-the-art technology not only brings the students closer to the real world but also kindles the spirits to innovate and excel over and beyond what already exists.
[pic]
Figure 3-1: Clock-speed of the Electronic Chip Industry
3.4.2 Technology TCO
Total cost of ownership is one of the biggest deterrents for educational institutions to acquiring technologies. The costs include not only the software and hardware, which is primarily available only at a premium in most developing environments because of the costs incurred in importing them as opposed to manufacturing or developing them in-house. The problem is two-fold, first the technology is not developed inside the country for the most part and second, there are very few incentives from the government for the local technology developers.
3.4.3 Technology Unavailability
Technology availability is a major concern among academia in developing environments because it is almost always imported from the developed world. The state-of-the-art technology is not only expensive; it starts to get attention in these environments only after an initial success in the developed world. Hence, these environments tend to become secondary or follow-on customers and never the lead adopters. Hence, they have little impact on what the technology should do, how it should look, or what it should cost.
3.4.4 Impediments to Innovation Adoption
There are several aspects to take into consideration when trying to deploy a technology-enabled solution for the issue at hand. Primarily, we cannot ignore the fact that many of the developing nations lack proper Information and Communication Technology (ICT) infrastructure. Some are going through recession and serious economic troubles and many do not have the required trained professionals to make the best use of any such assisting technology.
|Educational Infrastructure, India |Description of Decision Variables |
|Internet and Communications |On average, 1 in 50 colleges have internet access [76]. Many |
| |engineering colleges including the elite have a low-bandwidth |
| |connection to the Internet in the order of 2 Mbps at a maximum [77]. |
|Laboratories |Technologies used are 5 to 15 years older than the industry standard, |
| |setting aside the top few institutions. |
|Software |Software is either obsolete or pirated, which is a problem in itself. |
|Computer to Student Ratio |Roughly 1: 2000 [25] |
|Access to Research |Limited, local library is the primary source of information. |
|Access to experts in the field |Very limited access to experts who can guide the younger students and |
| |researchers in the field of their interest. |
|Policies | |
|Deterrents |Lack of strong policy on manufacturing electronic components, partial |
| |support from MNCs and lack of stringent rules on imports. |
|Propellants |Strong educational system, growth in the private equity and bank |
| |systems. |
Table 3-2: Technology Infrastructure in a Developing Environment
3.5 Populating the Academia-Industry Partnership Model
To populate the proposed model, key data on several variables are required. These variables are discussed in the following sub-sections.
3.5.1 Student Enrollment
Student enrollment continues to be on the study rise in universities and colleges across India, as presented in Table 3-3. Over the past decade the number of graduating students has doubled from roughly 4 million to 8 million. This is an important factor in the framework since it affects several aspects. First, increase in the number of students directly affects the student to computer ratio, which is extremely low today. Second, the number of students graduating has an affect on the job market. An increase in the number of qualified graduates certainly makes the multinational companies that want to outsource development for cheap labor. Third, the student perceived satisfaction is directly affected by the time he or she is allowed to work in the laboratories or take assistance from the faculty. The following table summarizes the growth of the educational sector in India from early 1950s to late 1990s.
|Institutions |1950-51 |1990-91 |2000-01 |
|Universities |30 |117 |254 |
|Colleges |750 |7346 |10200 |
|Enrolment('000s) |263 |4925 |7000 |
|Source: New Delhi: Ministry of Human Resources Development [78] |
Table 3-3: Higher Education Growth in India
3.5.2 Research Budget
For many academic institutions in India, research budget is an alien concept. The problem stems from the fact that there is little or no interaction between the academic institution and the industry or the government to attract funds for research. Even though there are extremely brilliant students enrolled in some of these institutions, their energy and aspirations soon wane down once they realize that there is not much scope for creativity and innovation.
3.5.3 Institution’s Innovation and Intellectual Property
One of the more successful models in academia in the developed world is the continuous research and innovation in various aspects including science and technology, process innovation and improvements, bio and Nano technologies and so on. These innovations are made possible by committed researchers, very supportive industry and government partners and energetic and extremely talented students. Innovations are the lifeblood of many leading educational institutions in the US, for example, MIT’s innovations have created numerous products, over 4000 companies, over $200 billion market value, and well over a million jobs. MIT derives licensing revenues in millions on the innovations through intellectual property right that are then funneled back into further research and advancement of the facilities at the institute.
3.5.4 Institutions’ Reputation
Educational institutions in India in general have very high reputation outside the country. Part of the reason is India’s qualified software professionals who have proved themselves as competitive engineers, entrepreneurs and leaders in the software field. However, there is very little that has been done by the industry and government inside the country that leverages the reputed academic institutions. Simple ideas such as branding and marketing the elite institutions in order to attract investments from multinational companies could go a long way in helping these institutions advance. Reputation of the institution springs from three different sources, its students’ performance in the industry, its researchers’ breakthroughs and its leaders’ vision.
3.5.5 Research Grants
In the United States, educational institutions receive grants from various sources such as the government agencies, industry partners, non-profit foundations, wealthy individuals, alumni and so on. In India, almost all the funding for public academic institutions comes from the central and or state governments. This is a striking difference not only in the sources of funding but also in the manner in which the system is structured. As discussed in the thesis outline section, there is a very loose coupling between the industry, government and the academia unlike here in the United States.
3.5.6 Academia-Industry-Government Value Chain
Academia, industry and the government are interdependent in order to function effectively. A hypothetical look at the value chain is shown in Figure 3-2. The value chain starts at the bottom with students and researchers who enroll in academic institutions to gain knowledge and thereby become qualified to pursue employment opportunity at a company. Next in the value chain is the academia, which provides training to the enrolled students, infrastructure support for enabling better learning and strong ties with the industry. On the other hand, academia also provides innovative solutions to problems faced by the industry, while in return it receives research grants and equipment grants from the industry. Industry benefits from the infrastructure support, governing policies and subsidies that the government provides and in return it generates taxes, growth in the economy and leadership through technology advancements. It is critical to understand this value chain to keep in perspective the importance of systems thinking, which is the right approach to solving the problem of digital divide in the academic institutions in developing environments.
[pic]
Figure 3-2: Academia-Industry-Government Value Chain
3.5.7 Recruiting
Industry hires graduates and resident students for jobs and internships. By bringing potential graduates in-house, companies benefit in terms of solving critical problems that need a fresh perspective. Industry will approach those institutions that are reputed, that have established themselves as experts in the field and whose students have consistently showed results. Hence, recruiting is an important variable to consider in designing a system’s solution framework.
Chapter IV assimilates the factors discussed thus far into a systems framework. The focus of the framework is to develop low-cost, flexible and synergistic solution that leverages advanced technologies and emphasizes industry, government and academia partnership.
Chapter IV: Systems Approach to the Solution
This section presents a systems approach to the digital divide issue in academia in the developing world. First, it outlines two real-world implementations of system-engineered solutions, analyzes how the solutions have fared since their deployment. The framework draws upon the successes and failures of these two implementations as well as key learning from projects discussed in Chapter 2, to build the proposed new systems framework. The proposed framework is structured as follows: First, an overview of the enabling technologies is presented. Second, a low-cost mobile wireless laboratory model is discussed. Third, design specification for a grid service that can run in the mobile wireless laboratory environment is presented. Fourth, a system dynamics model is created for the grid service and reference modes, sensitivity analysis, behavior patterns of some important aspects of the model and emergent dysfunctional properties are discussed. Finally, the business aspects of the framework including the industry and government support are discussed.
4.1 Real-world Implementation of System Engineered Solutions
In this section we will examine two real-world implementations of system engineered solutions and try to draw from their success and failure factors in developing the systems framework.
4.1.1 Leveraging Computer and Television Assets in Brazil to Deliver Educational Content
In a manner similar to broadcasting closed caption text on the TV, Gupta [6] and other researchers present a framework where Internet related material such as web page content could be broadcast from the TV station to its viewers by using the Vertical Blanking Interval (VBI). This provides a unique approach to solving the last mile communication infrastructure problem in developing environments through the efficient use of existing national assets such as TV signals and computers.
VBI consists of the first 25 lines in the PAL-M TV signal that are typically not used to transmit any information. In Brazil, which uses PAL-M signal, VBI provides a bandwidth of around 150 Kbps. During the hours when the TV station is not broadcasting regular TV material, the remaining 520 lines of signal could be used to broadcast data as well; this provides a total bandwidth of 4 Mbps.
The advantages of using this model are many; it leverages a well tested and established TV broadcast channel as the communication mechanism, and it piggybacks on the already existing broadcast signal and utilizes the unused spectrum, thereby increasing the efficiencies of the signal and it is scalable. The main drawback of this solution is that it requires VBI inserters and decoders. At least one inserter, which costs approximately US $5000, is needed for every station that wants to insert content into the TV signal. At the destination end, every destination device irrespective of whether it, a TV or a PC, must have a VBI decoder which costs US $100 to US $350 depending on the end device. Other drawbacks include: one way passive content flow, low bandwidth availability during the regular hours of broadcasting and limitation of the model to be extended beyond delivering content, such as online experimentation and problem solving.
4.1.2 Media Lab Asia
Media Lab Asia (MLA) [41] was conceived as an academic research program dedicated to bringing the benefits of new technologies to developing environments, with a special focus on meeting the grand challenges in learning, health, and economic development. The Government of India provided the seed funding of US $14 million, while expertise and the vision was driven by the parent organization, Media Lab, at MIT. MLA embarked on several projects in collaboration with the local elite research and academic institutions. Initially it appeared that the MLA had taken a systems approach to the problem of digital divide by bringing together the synergies among government, industry and academia.
Incorporated in Sep. 2001, MLA has four key focus areas for research; (i). Low-cost rural communication; (ii). Low-cost computing; (iii). Low-cost language interfaces and sensors; and (iv). Consolidation of the first three projects embodied in field research across the country.
MLA’s business model relies heavily on funds from the Government and from non-profit organizations. Several projects including the ones discussed in Chapter 2 have been created under these four focus areas. So far the impact of these projects on the developing economies has been negligible. Some consider the Media Lab Asia a failure [89, 93].
MLA was set up as a research and collaborative agreement between India and MIT. In April 2003, Department of Information Technology, India decided to restructure MLA to a project-based initiative, probably to incorporate accountability and well defined goals; the funding will now be provided entirely by the government. This is a major transition from a research organization mode. The new organization could be compared to a product development organization where the product, the target markets, the project goals, budget, deliverables and schedules are all well defined. In this kind of environment, unlike in the previous organizational structure at Media Lab Asia, the opportunity for variance, creativity and innovation is much smaller. The DIT has also further reduced the fund requirement of the project to US $200 million from the earlier US $600 million; the original vision was actually for US $1 billion.
There are several factors that contributed to the perception that MLA is a failure. First, the technologies developed so far have not had any significant positive impact on the rural communities of India and other developing environments and did not generate the critical mass necessary to develop economies of scale. Second, there appear to be open issues around the intellectual property rights [88]. It is not clearly defined who will be the beneficiaries of the innovations that emerge from MLA and to what extent. Third, there is problem of finances. With the overall global economy in a dire situation, funding for “next generation” research has become hard to find treasure. Finally, Media Lab has tried to bring together very diverse set of fields such as biology, engineering, art, music and so on to create unique products. However, the model has received wide criticism from the insiders as well [89], even for the main lab in the US.
4.2 Systems Thinking
Many organizations, groups and individuals have come to recognize the fact that the technology gap cannot be fixed by attacking the problem from any one angle. It is a system of systems issue and must be viewed from that perspective to devise an effective solution. The framework developed in this thesis focuses on the architectural aspects shown in the dotted box in Figure 4-1.
[pic]
Figure 4-1: Upstream and Downstream Influences in Architecting a System
The upstream influences such as low-cost technology alternatives, governing policies, industry interaction and requirement from the academia are evaluated. Downstream influences such as perceived quality of education, specific design tools and technologies, implementation costs, tactical and strategic feasibility and their effects on industry and academia partnership are evaluated.
4.3 Proposed Technology Framework
There are four primary goals behind the proposed framework.
Provide industry standard software and hardware solutions to students in colleges through collocated mobile laboratories.
Provide Internet connectivity through low-cost communication technologies in the last-mile or the school-mile.
Provide the software necessary to reserve resources, dynamically allocate unused bandwidth to applications.
Intelligently decide which network connection to use based on the quality of the link and other parameters.
There are several new advances in technology arena over the past decade that, when combined, are capable of providing an effective solution to the technology gap issue. These advances can be categorized into three complementary solution spaces as shown in the Figure 4-2. The first one is Internet and related technologies. Internet has become a household name in the United States, however, only now are the developing environments feeling its true impact. The number of Internet users in India, though very small (7 million), is growing very rapidly at a CAGR of 71% from 1998 to 2001. With the newer, less expensive technologies such as VFoIP, the affordability of telephone calls even across different continents goes up tremendously. Suddenly, students in a rural part of India have access to the best and brightest students, researchers and experts in the field from around the country as well as overseas. Students will become empowered with valuable course materials from renowned institutions like MIT through their Open Course Ware project.
Second, it is the wireless revolution, which has had positive effects on how remote areas are being ushered into the Internet era [4, 27]. Recent developments in the wireless standards, security [38], affordability, broadband and Internet have helped IEEE 802.1x technology to emerge as the strong leader in this pack. Already many enterprises, hotels, fast food chains and home users have been convinced by the power of mobility and its simplicity by being the lead users of this technology. Wi-Fi operates in the unlicensed spectrum 2.5 KHz to 5 KHz in the United States; however, in India Wi-Fi deployments are restricted to single campus networks. The government is working on opening up the spectrum for public use [17]. Finally, Grid computing is gaining momentum among many US and European institutions. Grid computing paradigm is based on the founding principle of sharing resources across multiple actual organizations to form virtual organizations that seamlessly allow applications, data, memory, disk space, computing power and network resources to be shared in a coordinated fashion. The kinds of applications Grid forum [57] is looking into are very high-end applications which require very large amounts of compute power, disk resources and network capacity that any one institution may not be able to possess. The Grid forum is developing recommended standards around sharing protocols, resource availability management and security and so on. This principle of sharing resources among organizations is an extraordinary match between the requirements for state-of-the-art technology laboratories and availability of idle resources in other partner institutions.
[pic]
Figure 4-2: Circles of Complementary Enabling Technologies
4.3.1 Internet Technologies
Virtual Private Networks (VPNs) [79] are being used as alternative to high cost leased line technology by enterprises and educational institutions alike. The main advantage of VPN technology is that instead of having a dedicated leased line between the source and the destination points, the Internet is used as the transport mechanism. Packets flowing between the source and the destination are encapsulated and sent in an encrypted fashion using an encryption algorithm such as PKI or DES forming a tunnel that shields VPN traffic from rest of the traffic.
4.3.2 Wireless Technologies
IEEE’s 802.1x protocol, also known as Wi-Fi, is gaining recognition as the solution to the last mile bottleneck in today’s telecom environment. The maximum bandwidth offered by 802.11a and 802.16 is 11 Mbps and the security protection is very weak. IEEE has been working on the security aspects and the bandwidth limitations that made their way into the 802.11g standard. 802.11g devices operate in the 5 KHz frequency. Wi-Fi Alliance is working on 802.11i and 802.1x that will use TKIP, AES security mechanisms [80]. Wireless technology is certainly the most effective, simple and affordable technology to provide connectivity to the students and researchers in the academia in developing environments.
Initially the Wi-Fi signal range was limited to a few hundred feet; however, more powerful high-gain antennas are coming into the market that can send and receive signals in the range of a few hundred kilometers. For example, recently, a Swedish company, named Alvarion, tested their high-gain antenna creating the record in Wi-Fi reach of 310 kilometers [30].
4.3.3 Grid Computing
Grid computing is an extension of campus-wide distributed computing architecture into a global network of compute resources, namely, processing capacity, storage capacity and network bandwidth, which enables coordinated sharing of these resources through standard protocols. The current focus of Grids is on high-end applications that require heavy availability of compute resources. However, this same technology can be also used very efficiently to provide shared resources to low-end applications as well. To quote the inventors of the Grid computing paradigm, “A computational grid is a hardware and software infrastructure that provides dependable, consistent, pervasive, and inexpensive access to high-end computational capabilities.” [58]
There are basically four types of resources that can be shared; they’re processing power, storage, applications and network. Grid infrastructure tries to re-use the IP and related protocols; thereby it not only avoids reinventing the wheel but also takes advantage of an already proven concept and a working system.
4.3.2.1 Grid Technology Overview
The Grid protocols, services and infrastructure can provide on-demand compute resources to even the most remote parts of the world, where there is a need for compute resources to perform low-end application processing, through lightweight application services developed on top of the Grid infrastructure.
[pic]
Figure 4-3: High-level Grid Architecture and Functional Blocks
Sun Microsystems describes these low-end applications as “Standard daily workloads consisting of many medium-size, single-threaded jobs, each running a few minutes to a few hours, without the need to provide interactive results. “ [60].
Figure 4-3 depicts the Grid architecture. There are striking similarities between the Grid stack and the Internet OSI model. It is a deliberate decision on designing the Grid architecture around the already widely successful and tested IP architecture. Grid architecture is first and foremost a protocol-based architecture, protocols define the mechanism by which VO users and resources negotiate, establish, manage, and exploit sharing relationships. A standards-based open architecture facilitates extensibility, interoperability, portability, and code sharing. Grid follows an hourglass model. Top of the hourglass is a broad spectrum of end-user applications, the middle portion representing the core of the Grid infrastructure, which is protocols and abstractions for resource identification, sharing and scheduling. The bottom portion interfaces with the low-level hardware.
4.3.2.2 How Grids can offer a Part of the Solution?
The basic principle underlying the Grid Computing concept is sharing of computing resources effectively across multiple research organizations in order to fulfill a common set of scientific and business goals. The key here is “sharing” resources. Today, most of the focus of researchers and grid deployments is around high-end applications such as genomics, visualization of medical records, collaborated system design such as aircraft manufacturing and so on which require vast amounts of storage space for the data analysis, tremendous processing power and high bandwidth network connections to retrieve and or store data collected from/to remote sites such as chemical reactors etc. However, the same Grid technology could be easily applied to the digital divide problem in the academic sector of many developing environments, here the problem being lack of resources to perform even the basic tasks such as state-of-the-art compilers to develop programming exercises as part of the coursework, or performing mathematical calculations in a spreadsheet, or preparing a presentation. Many of these applications are out of reach for students in most colleges in developing environments due to budget constraints. Here, by sharing these resources from the Grid, one could enable the educational institution to provide their students access to state-of-the-art technologies at very low cost.
4.3.4 Mobile Wireless Enabled Laboratories
Wireless technology has penetrated at a feverish pace in some of the developing environments. For example, in India, the number of wireless telephone subscribers has grown to over 40 million in a short span of 5 years while the wired infrastructure has not been able to penetrate to even half that level in over 50 years. Wireless technology can and should be promoted in the developing environments for Internet connectivity as well. A simple wired network installed at IIT, Kanpur cost the institution roughly US $425,000. As noted in previous chapters, the majority of the academic institutions in India do not have even a tenth of this cost allocated for technology spending. Furthermore, this majority do not have ties into the local industry, the multinational corporations and the government unlike few of the elite academic institutions in the country, which makes it extremely hard for them to acquire state-of-the-art technologies and upgrade them as and when necessary. This section evaluates an alternative approach to providing state-of-the-art laboratories at a comparatively low-cost for the institution.
Kumar [34] and others have demonstrated the power of low-cost mobile computer stations that have a computer with wireless Internet access stationed on an easily available locomotive. It is a standalone unit that runs on a power generator and travels across several rural regions of India to provide information resources to the rural farmers, such as weather forecasts, modern farming techniques and so on. On the other hand, organizations including TeNeT and others are actively taking advantage of the advances in long-range wireless technologies to provide low-cost solutions to remote parts of India as well as other developing environments. Academic institutions are well positioned to benefit from the synergies among various enabling technologies. Following sections present a low-cost, wireless enabled mobile laboratory shared among several local academic institutions in the range of 50 Km -350 Km radius.
The architecture for the proposed mobile wireless laboratories is illustrated in Figure 4-5.
Required Technology: 1 Wireless enabled router that supports VPN, 1 Wireless access point, 10-20 PCs, 5-10 Workstations and 1-2 Servers.
Technology
Mobile laboratories are not uncommon; many enterprises in the United States make use of mobile laboratories to demonstrate their state-of-the-art research and products to customers by showcasing the relevant work in a mobile station that travels the country. The technology necessary to implement mobile laboratories is mature now, with an Alvarion’s long range Wi-Fi access point and computers and workstations installed with wireless PCI cards available in the market. The Wi-Fi capability built into the access point and the PCI card does Ethernet to Wi-Fi and vice-versa translation at the layer 2. Internet Protocol (IP), which has become the de facto network layer protocol for the Internet, is used as the layer 3 protocol for Wi-Fi.
Bandwidth Requirements
The bandwidth provided by the 802.11b technology is in the order of 11 Mbps and 802.11a is 56 Mbps but the long-range Wi-Fi is still limited to less than 10 Mbps and with a potential to be much higher in the future. The mobile laboratory can be extended with VSAT connectivity option, which is a widely adopted last mile solution in the academia in India today.
Policies
Sharing resources is fundamental to the proposed model, which requires policies defined to delineate the usage of resources by individuals and individual institutions. Some key aspects such as who can use the resources, who can share the resources with other, what resources can be shared, to what level, who pays for what resources, accounting, billing and reporting the resource usage to relevant applications or individuals must all be configured as policies, preferably using XML schemas.
The mobile laboratory is envisaged as a shared technology among engineering institutions in a particular location in the radius of 10 to 100 Kilometers.
Costs
Average cost of wireless access points and Wi-Fi cards has fallen significantly over the past few years and it continues to fall further. Around 1998 the combined kit would cost over US $500, but at the time of this writing (April 2003), it is possible to get one for under US $125. The Wi-Fi PCI cards are about $50. High-end PCs cost about US $1250 to US $ 2000, workstations and server machines cost around US $6000. A more detailed cost model will be discussed for the entire mobile laboratory solution later in Chapter 5.
Implementation options
In Figure 4-4, each remote site can be viewed as a local educational institution which has signed on with the lab service provider. Each site has an option to either have a wireless access point installed in their premises or sign on for the mobile lab service.
[pic]
Figure 4-4: Prototype Architecture for a Wireless (802.16) Enabled Mobile Laboratory
Individual sites subscribe to the mobile lab service and are provided with connectivity and access to the Internet at scheduled times. A simple calendar software tool incorporated into the lab service could provide functionality for the subscribers to schedule lab service in advance. The proposed solution can be implemented in three different phases. First, a wireless enabled “school-mile” mobile computer laboratory can be established as a kiosk. Second, the laboratory can be VPN enabled. Finally, a grid service such as the XLab can be installed in the laboratory which will not only bring the state-of-the-art software and hardware platforms to the students from across the country but also from around the globe. The following section describes the details of each phase one level further.
4.2.5.1 Phase I
A mobile laboratory draws on some of the successful models from the past and applies them to the academic scenario. It builds on these models by introducing VPN based technology; resource sharing software and network smarts such as automatic bandwidth management based on the link quality, application request and other parameters. The academic institution interested in the service will need to sign up for a local mobile laboratory service from the provider. We can envision different viable business models for this service ranging from a regular flat monthly subscription fee to usage-based fee. The goal of this section is not to formulate the best business model; it is best left to the service provider.
4.2.5.2 Phase II
In this phase Virtual Private Network functionality will be introduced. VPNs enable multiple subscribers and multiple applications to share the network bandwidth as though each of them has a dedicated link. This is possible with the newer layer 2 Ethernet switching technology that is gaining popularity in the enterprise today. However, VPNs also introduce several issues that need to be addressed. First, there is the security and privacy issue, each individual site’s data traffic must be secured and protected from the other traffic streams sharing the network bandwidth. Second, it involves guaranteeing quality of service and different classes of service for different sites, as well as different applications in the same site. Finally, it needs additional software and configuration on the client side and at the central hub site.
4.2.5.3 Phase III
Grid services are still evolving; many standards are yet to be defined and not many low-end applications are being run on the grids today. As discussed in the previous sections, grids can be leveraged for high-end application services to the academic institutions. For target institutions running high-end applications has only been a distant dream, grids may very well be the ideal mechanism that makes this dream a reality. In the future incarnations of the grid services, running low-end applications, sharing software, disk space, network connections and processing power will bring rich set of functionality available through the mobile.
4.2.5.4 Inside the laboratory
The mobile laboratory shall be equipped with industry standard computers, workstations, and communication equipment such as the wireless hub, a router and a content filtering and firewall switch as shown in Figure 4-5. It will also include standard operating systems such as Linux and UNIX variants such as Solaris and key application software and productivity software such as OpenOffice’s spreadsheet, presentation packages.
|[pic] |
Figure 4-5: State-of-the-art Mobile Laboratory
The mobile laboratory will be evaluated more closely in Chapter 5, where the framework developed in this thesis is applied to Indian academic environment in specific. A cost model, comparison with an already implemented solution in one of the Indian elite institutions is performed to see what benefits this low cost, flexible and state-of-the-art mobile laboratory could bring to academic institutions.
4.3.5 Architecture Design for a Grid Service Running in Mobile Laboratory
The infrastructure for computation hardware and connectivity were presented in the previous section. This section details a software service that uses Grid technologies and extends the capabilities of the infrastructure presented in the previous section.
Figure 4-6: Lab Service System Model using Grid Infrastructure
Key aspects in terms of architecting the system such as who is the end user, who will provide the service and to whom [1] have been well thought out before diving into the design. Figure 4-6 shows one of these models at a functional block level.
The functional blocks can be delineated into three planes:
Grid Application Plane
Grid Service Plane and,
Grid Resource Plane
The application plane consists of several functional modules. However, they can be categorized into the front-end which is the Graphical User Interface (GUI) and the service API into the Grid Service Plane. The back-end, which is the intelligence of XLab core that glues together many functional blocks such as security, error handling, service level agreements, policies and so on. Finally, there is the communication server that seamlessly connects the application to the required resources on the Grid. This section will first describe some of the important core modules in the Grid architecture [59], which are part of the Grid Service and Grid Resource Planes, and then follow it with description of XLab grid service and its functional modules.
Grid Architecture for Resource Allocation (GARA): GARA acts as the central repository of resources and manages multiple distributed GRAM modules. GARA has a “big picture” view of resources across multiple domains, unlike GRAM which is very specific to a particular administrative and/or network domain.
Monitoring and Data Services (MDS): MDS provides access to static and dynamic information of resources. Basically, it contains the following components:
a. Grid Resource Information Service (GRIS): GRIS is the repository of local resource information derived from information providers.
b. Grid Index Information Service (GIIS): GIIS is the repository that contains indexes of resource information registered by the GRIS and other GIISs.
c. Resource Information Provider Service (RIPS): Resource information providers translate the properties and status of local resources to the format defined in the schema and propagates this information up to the GRAM.
d. MDS client: A search for resource information that you want in your grid environment is initially performed by the MDS client.
Grid Resource Allocation Manager (GRAM): Main components of GRAM are:
a. Resource Specification Language (RSL): RSL is the language used by the clients to submit a job. All job submission requests are described in RSL, including the executable file and condition on which it must be executed. You can specify, for example, the amount of memory needed to execute a job in a remote machine.
b. Global Access to Secondary Storage (GASS): GRAM uses GASS for providing the mechanism to transfer the output file from servers to clients. Some APIs are provided under the GSI protocol to furnish secure transfers. This mechanism is used by the globusrun command, gatekeeper, and job manager.
c. Dynamically-Updated Request Online Coallocator (DUROC): By using the DUROC mechanism, users are able to submit jobs to different job managers at different hosts or to different job managers at the same host.
The XLab Service application is a lightweight process that runs on the host computer. It connects to the Grid for accessing compute resources. There are several functional blocks that must be implemented in order to have a basic working model of the XLab service on Grids. We can categorize the XLab service functional blocks in the following way:
i. Policy Control
ii. Grid API
iii. User Interface
iv. XLab Core Engine
4.3.5.1 Policy Control
XLab Policy Server functional block, shown in Figure 4-7, contains the following modules.
Service Level Agreement (SLA): The services offered by the Grid to XLab are monitored, measured and recorded by this module. It provides an interface specification which describes what level of resource sharing occurs, who can access the resources, when the resources can be used. It handles the error scenarios and exception handling when the agreement is infringed upon by the XLab client. Resource allocation is the control of specific resources including, identification and specification of the resource, the amount of resource to be allocated, the day and time the resource is to be allocated, backup and fail-over strategies. SLAs are a way to provide guarantees on resource availability. Resource usage violation monitoring includes aspects such as, what to do when a "user" tries to use more than the amount allocated to him/her/it. Resource usage accounting can be done on a per user basis.
Authentication and Authorization: The positive identification of the user and the application, with possibly encrypted traffic streams and the rights for the authenticated traffic stream to utilize the specific resources.
Policies: This module consists of policy server, policy client typically the policy decision point (PDP), policy object specification language, and the communication protocol such as LDAP or COPS. Access lists, SLA related policies, and other objects can be stored in the Directory Server shown in Figure 4-7 [90].
[pic]
Figure 4-7: XLab Grid Service Policy Interactions
4.3.5.2 Grid API
The Programming Interface into Grid consists of the Resource Specification Language (RSL) and a messaging environment to act as the transport mechanism. RSL is a notation that is understood by the grid components when application requests access to resources. It is based on eXtensible Markup Language (XML) and can be easily extended to incorporate new kinds of resources and attributes to the resources. For example, current grid implementations only understand the CPU and disk storage as resources, one of the new resources that will eventually be added is the network bandwidth. Hence, network bandwidth related specifications need to be incorporated into the RSL in order for any application to access network resources available in the grid. The transport can be implemented as straight TCP/IP socket or could be abstracted one level further and implemented in Java Messaging Services (JMS) or could even be encapsulated into Simple Object Access Protocol (SOAP) and transported over in Hyper Text Transfer Protocol (HTTP). Similarly, depending on the transport mechanism, we could apply different levels of security that the transport protocol supports, such as Secure Sockets Layer (SSL) if using HTTP and MD-5 and DES if using JMS and so on.
4.3.5.3 XLab User Interface
This interface provides the user of XLab grid service to authenticate the person, submit jobs, schedule resources, and view reports generated by the application. It also enables the XLab administrator to configure policies that get translated into policy objects and are stored in the policy database. For example, policy objects may be defined in XML [28] as follows:
4.3.5.4 XLab Core Engine
An integral part of the XLab service is the core module that glues together all the other modules, coordinates interactions with the Grid API and performs the overall exception handling for the service. The core module incorporates the XML parser such as DOM or SAX and the routines needed to traverse the object hierarchy. It is responsible for any standard I/O processes such as recording reports, creating, sending and receiving messages, creating subscribers, queues, topics, connection factories if JMS is used as the transport protocol. Incorporating SOAP parser and HTTP bindings if SOAP/HTTP is used as the encapsulation and messaging mechanism.
The timeline diagram in Figure 4-8 represents the protocol messaging that happens between the XLab Service and the Grids.
[pic]
Figure 4-8: Protocol for the XLab Service Application to use Grid Services
4.3.5.5 XLab Grid Service Description
In the Grid world, everything is a service. Resource discovery is a service, requesting access to resources is a service and applications that attach to the Grid become custom services to the end-user. Hence, the application needs to provide a service description in Web Services Description Language (WSDL) [28], a language defined on the XML primitives. The grammar for a service description can be represented as described next. The serviceDataDescription element has the following non-normative grammar:
?
*
New grid services must describe their capabilities using the above grammar, for example, the XLab grid service could be described as follows:
................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related download
- 2011 school library standards curriculum frameworks ca
- independent reading assignment
- english language arts content standards 1997 curriculum
- lesson plan language arts 6 8 story themes
- pornography s past sexual publishing
- the last lesson by alphonse daudet www
- strategies for providing students and researchers in
- book reviews from our electronic products where to
- delaware state university