A Sustainable Business Model Approach for Grid Computing – And a Life Sciences Example

Scholz S, Breitner MH, Blaurock M

Institut für Wirtschaftsinformatik, Leibniz Universität Hannover, Königsworther Platz 1, D-30167 Hannover

Telematikplattform für medizinische Forschungsnetze (TMF) e.V., MediGRID-Projekt, Neustädtische Kirchstraße 6, D-10117 Berlin

stefan.scholz@tmf-ev.de, breitner@iwi.uni-hannover.de, mathias.blaurock@tmf-ev.de

Abstract: The amount and shared use of digitalized data in scientific and business communities have increased. Several software applications require higher comput-ing power and more sophisticated network solutions. Grid computing tackles these requirements by sharing and coordinating distributed resources dynamically. Sev-eral grid projects have passed the first hurdles and are becoming technologically stable thanks to a huge public funding effort. But public funding will not last for-ever and clear and sustainable perspectives are necessary. But little is being done to generate sustainable business models suitable for grid computing. We propose a business model reference framework based on e(lectronic)-business findings. A thorough adaptation of such a framework is necessary if grid ventures are to be successful – and turn a profit in the middle and long terms. In addition, a life sci-ence example is outlined: The business model reference framework is applied and special emphasis is placed on the crucial issue of revenue models.

1 Introduction

In many sectors the amount of digitalized information is increasing. It is more and more common to use information technology to facilitate collaborative tasks and to handle huge amounts of distributed data. New discoveries, like the ‘decompiling’ of the human genome in 2001, gave an additional push to IT-relevant fields like bioinformatics [KR04]. This increases the demand for computing power, network performance and systems capable of manage this data. Grid computing tackle these needs.

Grid computing infrastructures support the sharing and coordinated use of often large scale resources in a dynamic and distributive environment. These resources are generally provided by different institutions forming a virtual organization [Fo01, BV05]. It is one of the emerging research fields in several scientific communities and industries.

Entrepreneurs in several sectors might take advantage of the new technology. This is especially be true for the engineering environment, the financial, as well as the life sci-ences and pharmaceutical sector. In these sectors real markets exist and some corporate grid-like computing solutions are already in use.

1345

A considerable amount of mainly publicly funded scientific grid initiatives around the world first led to technologically quite stable solutions. But public funding will not last forever and a sustainable future perspective will be demanded. On the commercial side, several suppliers provide already easy access to computing as well as storing resources and deliver proven grid software solutions. In addition, potent internet companies like Amazon are developing services similar to those of common grid initiatives.

However, many grid projects lack a clear view how to remain viable in the long run. Little research is being done to develop sustainable business models in this field. Thereby business models help to understand the big picture of the potential market and address crucial business-related issues.

Research efforts labeled as ‘grid economics’ concentrate mainly on models for efficient resource allocation within grids, e.g. [BS04], [Bu05], [Yu05]. While this is important for the inner view of the grid infrastructure, to understand the big picture a comprehensive discussion on sustainable and customer-oriented business models must be initiated. The internet ‘.com’ breakdown at the beginning of the 2000s demonstrated what it meant to push a new technological approach without understanding the specific business and its environment entirely. In the year 2000 alone, 210 internet companies with about 15,000 employees were closed down worldwide [HF02].

Economic research analyzed e-business relatively thoroughly and developed conclusive business model reference frameworks for e-business. Since the idea of grid computing can be considered as a further development of the internet approach [Ge06], business model frameworks from the e-business world might be a right basis for a sustainable business model approach in the grid computing environment.

In the following, we present a business model reference framework for grid computing based on considerations developed for e-business solutions. Taking the example of medicine and life sciences we demonstrate how to apply main elements of the presented framework to a promising field for this technological approach. We look in more detail at two customer-oriented aspects: the utility of applications and services offered, and revenue models. We link both concepts in our example of medicine and life sciences.

2 Methodology and Approach

This study is mainly based on grounded theory, considering the main academic literature (’state of the art’) and a thorough market analysis – reviewing more than 20 grid projects in medicine and life sciences. Furthermore, we talked to several community experts in order to prove our considerations.

In the following, we spotlight different e-business model reference frameworks and, in a next step, we transfer the distilled expertise to a business model framework for grid computing. We then discuss the main elements in more detail. In the fifth section we apply two crucial elements of the model to grid computing activities in medicine and life sciences. The work concludes with suggestions for further research areas.

1346

3 Business Model Reference Frameworks for E-Business

The term ‘business model’ is widely used and refers to business activities. However, the specific meaning often varies from a single definition and ranges from business process descriptions via marketing strategies to complex market structure frameworks. The pro-posed ideas are usually not contradictory, but they focus on different levels of the busi-ness world. We classified these different levels hierarchically (Figure 1, based on [Os05]). The first level gives a definition of business models. The second level present reference frameworks to describe a business model by different partial models. The third level comprises types of business within the value network, which apply the reference frameworks. These three levels are conceptual. The last two levels describe the im-plementation of business models in the real world. In this article we touch the first level, focus on the second and present a business model reference framework applicable for grid computing.

We define ‘business model’ as a simplified and structured description of all parts of a business in order to evaluate its sustainability, following e.g. [Bi02b], [Ma02], [Wi01]. Business model reference frameworks aggregate the different crucial business aspects in a logical manner.

Several partial models and frameworks which structure business models in the e-business environment can be found in the academic literature, see, e.g., [AZ01], [HB06], [St01], [Wi01], and for an overview see, e.g., [Sc03]. Several aspects within the different frameworks are recurrent. The majority of research papers highlight the following as-pects: the added value generated from the offered applications and services; the descrip-tion of the value chain and the role and organization of the different participants; the financial flow – as there are revenues and costs; and the consideration of the market environment. Several authors point out that an overall strategy of the venture is needed as a separate element. Only some research works stress that the applied technology in-fluences the business model significantly and should be regarded as a separate aspect. While the vast majority of research outlines the competitive environment, only a few touch specific legal and environmental conditions.

Figure 1: Business model hierarchy, based on [Os05]

1347

4 A Business Model Reference Framework for Grid Computing

4.1 Business Model Reference Framework

Grid computing internalizes all the possibilities and caveats that the internet offers – but raises it to a new level of complexity.

E-business models can facilitate the definition of a business model refer-ence framework appropri-ate for grid computing. To this end, we analyzed different frameworks des-cribed above and distilled a business model ap-proach for grid comput-ing, comprised of six main elements (Figure 2):

Core elements: 1. Utility view: considers overall strategic aspects and the value added by services

offered to the user 2. Value creation view: analyzes the participants, structures and processes of the value

chain as well as the capabilities of the participants 3. Capital view: focuses on the revenue model, costs and financing aspects

Conditional elements: 1. Technology: considers opportunities and caveats of the grid technology 2. Legal and sector-specific conditions: legal and sector specific environments 3. Market (competition and trends): considers market situation and dynamics The first three elements determine the core aspects of each business model and answer the crucial questions: What and who is my venture for? and what is the added value of the applications and services offered (utility view)? How do I deliver the value (value creation view)? And the essential question, do I gain money by running the business (capital view)? The latter points are the influencing frame for setting up the business. We call them the conditional elements. The grid technology determines the scope of possible business ventures, and a stable and trustful technology is one key to attract potential customers. The idea of distributing huge amounts of data within and beyond organiza-tional and political borders raises questions of legal and security issues that are until now unusual in the e-business world. Further, grid computing is in a very early stage of the product/application life cycle. The ’run to market’ is still ongoing and many grid tech-nologies and initiatives compete with each other.

Figure 2: Business model approach for grid computing

1348

The business model as a whole is a bird’s eye view of all analyses pertaining to one of the above-mentioned elements. The model proposed here was designed within the Ger-man life sciences and medicine MediGRID project1 [Vo07], [We07]. However, the ele-ments of this reference framework are generic for all grid computing ventures that pre-tend to gain market acceptance.

4.2 Utility View

The utility view describes the overall long term objective of a venture and defines:

- Who will be the customers, i.e. target groups?

- Which applications or services should be offered to them?

- What is the added value, i.e. utility, that a customer gains out of the applications and services offered or a partner of the value chain out of the collaboration?

The latter item points out that an additional perceived value is important not only for the final customer but also for all participants within the value chain [St01]. This is espe-cially true for a grid infrastructure. Grids comprise a set of individual partners that col-laborate in order to achieve a common goal. These partners form a ‘virtual’ community, held together only by a set of contracts. Only if all partners see a benefit the whole ven-ture can it be run successfully. Despite that fact, first business activities have to focus on the final customers and their acceptance of the applications and services offered by a grid infrastructure. If this test fails already, potential benefits or business models at dif-ferent levels within the value network are obsolete.

In order to formalize the utility of an application or service we introduce utility categories. Utility categories define the sources of added value for a certain product cate-gory. However, their relevance differs from product to product within one product category (Figure 3).

1 MediGRID forms part of the German grid initiative D-Grid [Dg07], funded by the German Federal Ministry of Education and Research.

Figure 3: Utility categories and perceived value

1349

The added value of a vehicle, for example, consists of the utility categories status, per-formance safety etc. Nevertheless, the relevance of each utility category is different, e.g. comparing SUVs and compact cars.

Utility categories help to define the best revenue model for a given product or applica-tion. Based on them, the final price a customer is willing to pay for a product or applica-tion can be deduced.

At this stage of our research we define four utility categories for grid computing:

- Computing: High-performance computing power, solving problems much faster

- Information: Access to high-quality data and information

- Algorithm: Provision of a unique application or middleware architecture

- Network: Fast and stable network connection to facilitate distributed collaboration.

These utility categories have to be revised in further studies, but are sufficient at this stage in order to demonstrate the basic systematic idea. These categories can be applied to reveal both the final willingness to pay and the best way to skim it.

4.3 Value Creation View

A business model should define what the architecture of the value creation looks like. This includes the different levels of the value chain or network and describes the differ-ent participants and their role within the value creation process [St01] – see a promising approach for grid computing in [Al07]. In addition, it has to stress the capabilities an organization dispose of and the one it needs in order to run the business successfully.

Grid computing raises the value creation process to a new level of complexity. The value network is represented by a dynamic virtual organization, glued by contracts, the Service Level Agreements (SLAs). SLAs define the details of the collaboration. Thus, workflows and collaboration are as good as these contracts are designed. This includes, among others, the clear definition of Quality of Services (QoS). Despite that fact, the organization relies on trust and commitment of the participants.

From an economic point of view, architectures that allocate resources efficiently and accomplish the clearing of cost between the different participants are needed. Promising approaches have been developed within the recent years, see, e.g., [Ar06], [Bu05].

Grid initiatives have to decide which services to ‘make‘ and which services to buy-in. Different commercial IT-vendors provide already easy access to computing as well as storing resources and deliver stable grid software solutions.

The value creation process within grid infrastructures can not be covered entirely within this article. However, the highlighted aspects give a first glance of what grid computing faces look like in order to run a consistent process of value creation.

1350

4.4 Capital View and Revenue Models

The capital view answers the important question, will the venture be sustainable in the long run, that is, will it turn a profit? It comprises the revenue model and the cost model. In addition it asks for a convincing model to finance the venture – especially in the start-ing phase of a business.

Profit means that revenues exceed cost. Since cost is to a certain extent determined by the business idea, the crucial question is whether somebody is willing to pay for the offered application or service. Thus, revenue models become the focal point of interest.

We believe that a revenue model consists of the revenue source, the revenue partner and a price scheme, following [Sk05], influenced by the utility, i.e. the added value gener-ated by the revenue source (Figure 4).

The revenue source indi-cates what will be priced. In most cases, it is the application itself. How-ever, other revenue sources are possible: information (i.e. data-mining revenues), contact (to the user or others, e.g. by advertising placement) or public value (e.g. through reduced costs in public health services). In the following, we will focus on the ’product’ represented by different application and services offered within a grid. The revenue partner indicates who will be paying. The price scheme determines possible pricing mechanisms. Relevant price schemes are manifold and several frame-works can be found, e.g. [BH05], [Wi01]. We have focused on the price schemes pre-sented in Table 1.

Transaction dependent

– Full dynamic pay for use, based on capacity usage (intensity driven) – Level pay for use, e.g. day/night or intensity driven – Fixed direct price per use, i.e. frequency driven

Non-transaction dependent

– Membership (regular flat rate) – Subscription (payment once)

Table 1: Price schemes

As stated before, utility categories indicate what the overall value added of a application or service is, and what can finally be priced, i.e. for which specific feature a potential customer is willing to pay.

Figure 4: Revenue model for grid computing

1351

Here we focus on the latter aspect and demonstrate how revenue models can be defined. To that end, we matched the utility categories with specific price schemes. The relation-ship between the different utility categories and possible price schemes is summarized in Table 3.

Computing Information Algorithm Network

DPR (CU high), LPR (CU low)

MS, FPR MS, FPR LPR, MS, FPR

DPR – dynamic direct pricing, LPR – level direct pricing, FPR – fixed direct pricing, MS – membership fee; CU – capacity utilization

Table 3: Price schemes for utility categories

High-performance computing power is resource-expensive, and resource scarcity is common in well-utilized grids. Prices should reflect this scarcity and must therefore be dynamic (e.g. by using auction models). If moderate computing must result in lower fluctuation, level pricing is more appropriate.

In contrast, information value is generally stable. It depends on the behavior of the reve-nue partner (frequent or sporadic usage) - if a fixed pay-per-use price or a membership fee should be applied. Algorithms are unique corporate knowledge that should not be given away, and therefore priced by regular payments as fixed pay-per-use or member-ship fee. Network capacity, i.e. bandwidth, might be scarce at a certain point. Thus, pricing depends strongly on the level of scarcity – fixed price schemes and level price schemes may both be applicable. Subscriptions might be useful to support certain activi-ties. But since revenues will be received only once, they should always be combined with other price schemes and never used alone.

5 Business Models for Healthgrids – How to Price Grid Activities

In the following we apply the prior ideas for utility categories and revenues models to the healthgrid community and demonstrate how to price grid activities. For that purpose, we studied the activities of 23 worldwide healthgrid projects. They include the promis-ing European projects MediGRID [Vo07], [We07], MammoGrid/MammGrid+ [Mc06] and Wisdom [Wi07]. The majority of the analyzed projects still lack a conclusive reve-nue model.

In a first step, we clustered the activities in seven application groups. We based our clus-tering mainly on qualitative criteria that enable a mutual exclusive separation of the activities, e.g. needed computing power or needed ad hoc accessibility.

However, several activities or applications are especially important for the healthgrid community but do not fit exactly in a mutual exclusive framework, e.g. activities related to digital images. Therefore, at this stage of our research, the clusters consider both the defined criteria and the distinctive importance of specific applications.

1352

Computationally expensive proc-esses:

These activities are very computationally expensive, usually planned in advance, and not needed ad hoc; in many cases they take advantage of a unique algorithm. These include, e.g., epidemiological studies, biomedi-cal calculations and simulations, pharmaceutical IT-based research.

Visualization and modeling:

These activities are computationally expensive, usually needed in a short- to mid-range timeframe and take advantage of a unique algorithm. They include therapy planning and visualization of research activities.

Digital image processing:

These activities are computationally expensive, often needed ad hoc or in a shorter time frame, and take advantage of a unique algorithm. They include diverse digital image diagnostic processes.

Image handling:

These activities rely on storage capacity, might be bandwidth expensive and take advantage of sophisticated data virtualization algorithms.

Data gathering and handling:

These activities need large amounts of storage capacity, take advantage of sophisticated data virtualization algorithms and rely on efficient inte-gration into clinical processes. These activities include clinical research activities and patient data managing.

Information data-bases:

These activities rely on large storage capacity and – depending on the final activity – close integration into other applications.

Real time medical care provision:

Depending on the degree of sophistication, these activities need regular computing capacities, but rely on a constant, stable and fast network connection with guaranteed bandwidth and generally on ad-hoc availabil-ity at any time.

Table 5: Grid computing application groups in medicine and life sciences

In a next step, we match the identified application groups with the prior defined utility categories as seen in Table 6. We analyze the level of utility for each potential revenue partner for every application group. In the life sciences, revenue partners are the life sciences industry, researchers, (clinical) specialists, physicians, and patients. The latter are not regarded at this stage of investigation. The results are given in Table 8.

Utility category Application group Computing Information Algorithm Network

Simulations/calc. ++ + Visualization/m. + ++ Image processing + ++ + Image handling ++ ++ + Data gathering/ h. ++ ++ Databases ++ Medical care prov. + + ++ ++ high utility; + low utility

Table 6: Relevance of utility categories per application group

1353

Revenue partner Application group Industry Researchers Specialists Physicians

Simulations/calc. ++ ++ Visualization/m. ++ ++ + Image processing + ++ ++ + Image handling + ++ ++ ++ Data gathering/ h. + ++ ++ + Databases ++ + Medical care prov. ++

++ high utility (frequent use) + low utility (sporadic use)

Table 8: Utility of revenue partner per application group

In a final step, we merge all our results as seen in Table 10. We can derive two main conclusions from the last table. First, every combination of revenue partner and applica-tion group will have its unique revenue model. Second, certain activity categories are only interesting for a narrow group of revenue partners. Thus, before pursuing concrete pricing models, grid initiatives need to define their activities and customers clearly. They have to identify the concrete value that each particular application promises to the users.

Revenue partner Application group Industry Researchers Specialists Physicians

Simulations/calc. DPR DPR Visualization/m. LPR

MS LPR MS

LPR Image processing LPR LPR

MS LPR MS

LPR Image handling FPR MS MS MS Data gathering/ h. FPR MS MS FPR Databases MS FPR Medical care prov. LPR

MS

DPR – dynamic direct pricing, LPR – level direct pricing, FPR – fixed direct pricing, MS – membership fee

Table 10: Generic revenue models for healthgrid application groups

6 Conclusions and Outlook

Grid computing is maturing and entering a stage where economic objectives come to the fore. From a business point of view it is crucial to draft the big picture before targeting the details. Business model reference frameworks, when adapted to grid computing re-quirements, help to face this challenge. The experience of the World Wide Web and e-business facilitates the development of such frameworks.

Based on these frameworks grid entrepreneurs learn to understand the final market and the needed resources and capabilities. On the market side often unanswered questions arise: Who will be the final customer? What specific needs do they have? What size is the market, what is the revenue? Finally, entrepreneurs have to prove whether potential users are willing to pay a sufficient price for applications and services to run a sustain-able business.

1354

As soon as these questions are answered, grid providers can concentrate on the resources available and needed to provide appropriate services. Hurdles with regard to the ’orches-tration’ of a virtual organization have to be solved. A convincing approach is needed for how to manage the virtual organization efficiently. This includes elaborated Service Level Agreements. Resource allocation and accounting models are already available. However, these models generally lack a real world business check. Another problem arises from the issue of software licensing. If services and application will be shared throughout a distributed environment, new licensing models will be needed.

In addition, community and application-specific conditions and restrictions have to be considered. In life sciences, data protection and security are crucial issues that have yet to be solved satisfactorily. Different legal systems might restrict ’across border ap-proaches’ for distributing resources.

Taking the example of life sciences and medicine, we demonstrate how to approach the market side and how to identify potential sources of added value. Finally we propose a framework for how to address the difficult topic of revenue models. Further studies are needed to enhance and validate the proposed framework.

Reference frameworks are strongly needed in order to understand the big picture of a business. Nevertheless they are still abstract models. Researchers should walk down the business model hierarchy introduced here, and identify viable types of business models for grid computing within the value network. They should also consider existing niche vendors of commercial grid solutions and understand their business models.

Grid computing offers attractive opportunities based on a new technology. But while pushing the mentioned topics forward, we should never forget one thing: customers do not ask for a technology – customers want us to solve a specific problem or satisfy a certain need, i.e. customers ask for a relevant service.

Acknowledgements

We especially thank the members of the MediGRID project, in particular Prof. Dr. O. Rienhoff, Prof. Dr. U. Sax, PD Dr. A. Weisbecker and Dr. D. Krefting. We also like to thank H. Schmidt from JSW Consulting for discussing business-related issues.

References

[AZ01] Alt, R.; Zimmermann, H.-D.: Preface: Introduction to Special Section - Business Mod-els. Electronic Markets 11, 2001; pp. 3-9.

[Al07] Altmann, J.; Ion, M.; Bany Mohammed, A. A.: Taxonomy of Grid Business Models. GECON 2007, Springer LNCS, Rennes, France, 2007.

[Ar06] Ardaiz, O., Artigas, P.; Eymann, T.; Freitag, F.; Navarro, L.; Reinicke, M.: The Catal-laxy Approach for Decentralized Economic-based Allocation in Grid Resource and Serv-ice Markets. Applied Intelligence 25, 2006; pp. 131-145.

1355

[BS04] Bhargava, H. K.; Sundaresan, S.: Computing as Utility: Managing Availability, Com-mitment, and Pricing Through Contingent Bid Auctions. Journal of Management Information Systems 21, 2004; pp. 201-227.

[Bi02] Bieger, T.; Bickhoff, N.; z. Knyphausen-Aufseß, N.: Geschäftsmodelle - eine neue Per-spektive der Managementlehre. In (Bieger, T.; Bickhoff, N.; Caspers, R.; z. Knyphausen-Aufseß, D.; Reding, K.eds.): Zukünftige Geschäftsmodelle: Konzept und Anwendung in der Netzökonomie. Springer, Berlin, Heidelberg, New York, Barcelona, Hongkong, Lon-don, Mailand, Paris, Tokio, 2002; p. 3-14 (German).

[BH05] Breitner, M. H.; Hoppe, G.: A Glimpse at Business Models and Evaluation Approaches for E-Learning. In (Breitner, M. H.; Hoppe, G. eds.): E-Learning: Einsatzkonzepte und Geschäftsmodelle. Springer/Physica-Verlag, Heidelberg, 2005; pp. 179-193.

[Bu05] Buyya, R.; Abramson, D.; Venugopal, S.: The Grid Economy. Proceedings of the IEEE 93, 2005; pp. 698-714.

[BV05] Buyya, R.; Venugopal, S.: A Gentle Introduction to Grid Computing and Technologies. CSI Communications, Vol.29, No.1, Computer Society of India (CSI), 2005; pp. 9-19.

[Dg07] D-Grid initiative. hhtp://www.d-grid.de, last visited 07/09/23. [Em07] Empirica: Registry of HealthGrid Implementations Across the EU and Beyond. 2007. [Fo01] Foster, I.; Kesselman, C.; Tuecke, S.: The Anatomy of the Grid. International Journal of

High Performance Computing Applications 15, 2001; p. 200. [Ge06] Geiger, A.: Service Grids – von der Vision zur Realität. In (Barth, T. Schüll, A. eds.):

Grid Computing: Konzepte – Technologien - Anwendungen. Vieweg, Wiesbaden, 2006; p. 17-32 (German).

[HB04] Hoppe, G. Breitner, M. H.: Evaluation and Optimization of E-Learning Scenarios. Zeitschrift für Betriebswirtschaft (ZfB), Spec. Iss. 2/06, 2006; pp. 43-62.

[HF02] Hirakubo N, Friedman HH.: Dot-Bombs: Lessons from the Dot-Com Debacle. Journal of Internet Commerce 1 (2), 2002, pp. 89-102.

[KR04] Kesh, S.; Raghpathi, W.: Critical Issues in Bioinformatics and Computing. Perspectives in Health Information Management, Vol. 1, 2004; 1:9.

[Ma02] Magretta, J.: Why Business Models Matter. Harvard Business Review 80(5), 2002; pp. 86-92.

[Mc06] McClatchey, R. H.; Manset, D.; Solomonides, A. E.: Lessons Learned from MammoGrid for Integrated Biomedical Solutions. In (IEEE Computer Society eds.): 19th IEEE Sym-posium on Computer-Based Medical Systems, CBMS'06, 2006; p. 745.

[Os05] Osterwalder, A.; Pigneur, Y.; Tucci, C. L.: Clarifying Business Models: Origins, Present and Future of the Concept. CAIS 16, 2005; pp. 1-25.

[Sc03] Scheer, C.; Deelmann, T.; Loos, P.: Geschäftsmodelle und internetbasierte Geschäfts-modelle-Begriffsbestimmung und Teilnehmermodell. ISYM Working Paper, http://isym.bwl.uni-mainz.de/publikationen/isym012.pdf, last visited 07/09/23, 2003.

[Sk05] Skiera, B.; Spann, M.; Walz, U.: Erlösquellen und Preismodelle für den Business-to-Consumer Bereich im Internet. Wirtschaftsinformatik 47, 2005; pp. 285-294 (German).

[St01] Stähler, P.: Geschäftsmodelle in der digitalen Ökonomie: Merkmale, Strategien und Auswirkungen. Josef Eul Verlag, Köln-Lohmar, 2001 (German).

[Vo07] Vossberg, M.; Krefting, D.; Tolxdorff, T.: Grid Computing in der Medizinischen Bild-verarbeitung: Das MediGRID-Projekt. Proceedings des Workshops Bildverarbeitung für die Medizin 2007, Algorithmen-Systeme-Anwendungen, 2007 (German).

[We07] Weisbecker, A.: Grid Computing für Medizin und Lebenswissenschaften. In: Jahrbuch der Gesundheitswirtschaft 2007: Modernisierung, Prozessoptimierung und Vernetzung im deutschen Gesundheitswesen, Wegweiser, Berlin, 2007; pp. 166-167 (German).

[Wi01] Wirtz, B. W.: Electronic Business. Gabler, Wiesbaden, 2001; pp. 210-218 (German). [Wi07] Wisdom Project. http://wisdom.eu-egee.fr/, last visited 07/09/25. [Yu05] Yu, J.; Li, M.; Hong, F.: Resource Allocation Based on Pricing for Grid Computing

Environments. Lecture Notes in Computer Science, 2005; pp. 506-513.

1356