For exploring the quality features we will design two reference scenarios and realise corresponding prototypes. The reference scenarios will differentiate in their basic quality standards.

Reference scenario A: A P2P Based Communication System for Disaster Relief Operations
By 'disaster' we mean the actual deployment of personnel and equipment. the effective coordination of various ressources, for example the fire department starting with the different brigades that will have to be coordinated up to major incidents such as natural disasters, catastrophes caused by industries, terrorist attacks and war aftermaths. The scenario focuses on the immediate rescue efforts within the "first hours".

Three deliberations or developments suggest that features of P2P systems and mechanisms will reappear in reference scenario A:

  1. The always new arrangement of the involved organisations (influenced by the location of the disaster, vicinity and possible time of arrival of rescue teams), depending on severity and pecularities (i. e. required expertise) also varying in number and professional composition.
    These are the features of highly dynamic application configurations that are very similar to conventional P2P systems.
  2. During such events the communication networks often become impaired; universally applicable IT solutions should therefore be able to manage without a fixed line network infrastructure (ad hoc connection) but also use an increasingly improved infrastructure as best as possible. These, too, are features of P2P systems, because they are meant to operate without a "server" and without the reliability of partners being available.

This applies to all the quality features suggested by the research group, even if the impact is not the same in each conceivable scenario. The following characteristics directly entail high quality requirements: realtime conditions, highly dynamic node-/connection-topology, also application-wise high dynamics with regard to information sources and information processing elements, associated subsystems (rescue organisations) as well as objectives.

Reference Scenario B: The Integration of Software Development Tools into P2P Technology

Nowadays often hundreds to thousands of persons are involved in large software development projects. These people usually work at different locations all over the world and in different shifts "around the clock" and belong to different organisations. Coordinating the various sub-activities and maintaining the consistency of the resulting documents (in their varying versions and variants) is one of the main issues.


State of the art support for distributed development teams is usually and with almost no exception based on centralized servers. Such development platforms, however, have limitations when processing large commercial projects:

  1. Often the integrated solutions are not scalable and can barely be used for coordinating the activities of many hundreds of employees that not only "migrate" and are located all over the world but that also have very different connections to central servers.
  2. Tools necessary to process certain documents are available only at selected locations and can for several reasons (both technical as well as legal) not be used unlimited in a network spanning several companies.
  3. Normally, involved enterprises are also not willing (up to a certain point in time) to make confidential documents available on servers outside of their own intranet. Instead they will allow specific partners to view the development documentation at a certain, pre-defined point in time.
  4. Beyond that it is not acceptable for an organisation, also for reasons of availability, to store documents crucial for its own development processes (only) on servers administrated by a third party.

The hope is to achieve with P2P solutions a clear reduction in administrative expenditures, a more efficient usage of bandwidths for the exchange of development documentation, a clear reduction of the central operating resources (servers) to be provided and a decreased susceptibility with regard to malfunctions in single connections and nodes.

All in all this reference scenario is deliberately quite challenging for the P2P approach, because systems based on central servers have inherent strengths where consistent and reliable data administration is concerned. This scenario is therefore especially suited as a reference scenario, in particular for identifying the limitations of the P2P paradigm.

In contrast to the previous reference scenario this scenario puts higher demands on the quality features validity and trustability. All data generated in a project will most likely have to be kept available and valid within a P2P software development environment at first. Beyond that changes in data have to be retraceable and revocable. Protecting the data against unauthorized access and modifications on harmful nodes is also one of the key roles.

But there will also be new challenges for the quality features adaptivity and efficiency that do not so much result from the number of the peers and the dynamics of the considering networks, but instead from the high number of the data objects to be administrated (documents with an internal structur) together with their associated relationships and integrity conditions.

Contact

Prof. Dr.-Ing. Ralf Steinmetz


Fachgebiet Multimedia Kommunikation
Rundeturmstr. 10
64283 Darmstadt
S3/20

+49 6151 16-6150

A A A | Print | Legal note | Sitemap | Search | Contact
to topto top