MMAPPS software release

2nd MMAPPS Day Event

MMAPPS Introduction & Overview v2.5 (pdf)

MMAPPS Newsletter

Previous Project: M3I


Private Server

MMAPPS - Market Management of Peer to Peer Services

The MMAPPS project is researching how to use techniques from economics and social science to tackle some of the fundamental difficulties in creating well-founded, and therefore sustainable, P2P applications. The project's central approach is to extend techniques such as market management so that cooperation between peers can be encouraged without damaging the community-oriented structure of P2P architectures.

Existing P2P applications often work adequately when the cost of contribution is low and the value of consumption is high. But, in general, they have a problem in providing appropriate incentives to the peers involved. Each peer asks themselves "Why should I contribute (at some cost to myself), when I can still consume without it?" Many P2P applications function on the altruistic contributions of a small minority of peers with all the other peers "free-riding". The utility, efficiency, and robustness of such applications could be hugely improved by providing incentives for peers to contribute their resources, so that it becomes rational for individual peers to choose to contribute. The standard suggestion for providing incentives has been to use micropayments for transactions. However, payment schemes can often act against the collaborative, community-forming nature of P2P. So MMAPPS is looking to enhance such schemes with more innovative, non payment-based accounting schemes such as ratings, where individual members of a particular community receive a rating score based upon their contribution. This rating then affects how other members provide services to that individual.

The project is drawing upon current economic theory and social studies (partly conducted from within our project) to develop a generic toolkit that provides a collection of such incentive schemes, suitable for supporting a wide range of P2P applications. The toolkit will support application developers in devising a scheme tailored to their own application and community. This scheme can use payment where appropriate but alongside other, community-based, accounting, reward and punishment systems. The project is also developing the descriptive and analytical tools that can be used to study the viability and appropriateness of such schemes.

Among the application scenarios being investigated are:

  • Filesharing: As the prototypical P2P application, we believe that investigating this scenario will allow our results to be compared with existing work, and thus make our paradigm more immediately attractive to the current-generation of P2P application developers.
  • Peering of WLAN domains: in which local WLAN administrators enable connectivity to roaming nodes in exchange for future connectivity on their local WLAN network.
  • Semi-automated selection of telemedicine consultants: in which physically-remote communities are able to receive rapid diagnosis from a small pool (virtual team) of trusted experts.

Milestones

Major milestones are planned to coincide with the External Reviews of the Project:

Review 1 (6 months) requirements and basic draft of architecture definition finalised;
Review 2 (12 months) architecture completed, software modules specified, economic models created;
Review 3 (18 months) modelling complete, trial system built and trials initiated;
Review 4 (24 months) trial system completed and feedback into system's architecture and modelling. Trial and experiment definitions completed;
Final Review (30 months) Results of trials and experiments published, along with all final deliverables.

For Further Information

Please contact the Webmaster, the Project Manager or the Technical Director of the project.


© Copyright 2004, the Members of the MMAPPS Consortium
Site last updated:  08 Dec 2004