Objective
Traditional centralised system architectures are ever more inadequate. A good understanding is lacking of future decentralised peer-to-peer (P2P) models for collaboration and computing, both of how to build them robustly and of what can be built. The PEPITO project will investigate completely decentralised models of P2P computing.
It will:
(1) study the use-models of P2P systems, that is how they are perceived by users and what new applications are possible;
(2) develop the foundations of P2P computing, including formal foundations (calculi, proof techniques, security and resource models) and new distributed algorithms (for diffusing information and coping with multi-consistent views);
(3) provide a language-independent distribution subsystem tailored for P2P computing and;
(4) provide programming languages and platforms using this, showing that they are useful by implementing convincing demonstrator applications. Traditional centralised system architectures are ever more inadequate. A good understanding is lacking of future decentralised peer-to-peer (P2P) models for collaboration and computing, both of how to build them robustly and of what can be built. The PEPITO project will investigate completely decentralised models of P2P computing.
It will:
(1) study the use-models of P2P systems, that is how they are perceived by users and what new applications are possible;
(2) develop the foundations of P2P computing, including formal foundations (calculi, proof techniques, security and resource models) and new distributed algorithms (for diffusing information and coping with multi-consistent views);
(3) provide a language-independent distribution subsystem tailored for P2P computing and;
(4) provide programming languages and platforms using this, showing that they are useful by implementing convincing demonstrator applications.
OBJECTIVES
Peer-to-peer computing (P2P) is a paradigm in which applications are connected to a shared network as peers, that is with the same capabilities and responsibilities. Current P2P applications are limited to information exchange.
The objectives are to remove this limitation by:
- developing formal models to understand P2P computing;
- developing the distributed algorithms required for implementation;
- implementing a language-independent set of basic services;
- implementing languages and devise programming techniques and convincing demonstrator applications.
Further objectives are:
- better using resources at the network's edge;
- scaling better than server-centric computing;
- allowing device mobility (independence of IP address);
- allowing individuals to publish information and services, and allowing individuals to collaborate while remaining anonymous.
DESCRIPTION OF WORK
PEPITO will assume a completely decentralised architecture in which a peer can have four simultaneous roles: it may use services, provide services, forward requests, and provide caching of information. We also assume that peer nodes connect through a virtual network that is dynamic and intermittent, and that nodes do not possess a fixed IP address. To successfully deal with the complexity of P2P systems (in which failure, reconfiguration and security are central) it is important to pursue use-model analysis, theoretical work and prototyping in a closely linked style. The complementary expertise of the PEPITO partners makes this possible: the objectives will be addressed, but enabling interaction between them is also crucial. Use-model analysis of this type of system will investigate how they are perceived by users, and what new applications are possible.
Theoretical work will study the foundational concepts of P2P systems. This includes mathematical models (calculi, proof techniques, security and resource models) and new distributed algorithms (decentralised algorithms for diffusing information, and for coping with multi-consistent computing - with simultaneous inconsistent views of entities). System Design and Prototyping will develop prototypes of programming languages and programming platforms (middleware) suitable for peer-to-peer computing (such platforms are lacking today; those existing are server-centric).
One aspect will be a scalable and robust name/directory service based on our algorithms.
Together, all these will enable development of applications that:
- handle dynamic connectivity and device mobility;
- allow individuals to become publishers of information and services;
- permit full use of existing network resources at the edge of the network;
- and permit applications to scale better than server-centric designs.
Fields of science
Call for proposal
Data not availableFunding Scheme
CSC - Cost-sharing contractsCoordinator
164 29 KISTA
Sweden