The World of Peer-to-Peer (P2P)/New models

"New" models

Fault-Tolerant Web Sites

Many people have speculated that peer-to-peer file sharing technology could be used to improve wiki and other kinds of Internet services.

High quality video or large files distribution

The Internet infrastructure was not designed to support broadcasting. P2P partially solves this infrastructural bottleneck by switching the server or content provider from a single point to a decentralized infrastructure, that depends not on the specific network limitations but on the protocol that optimizes the distribution and its popularity.

In February 2008 the European Union announced its commitment into a four-year project that aims to create an open source, peer-to-peer BitTorrent-like client called P2P-Next, based on an improvement of the Delft University of Technology python project Tribler. The EU will contribute 14 million euros (£10.5 million, $22 million) into this project and another 5 million euros (£3.7 million, $7.4 million) will be added by another 21 partners that includes the European Broadcasting Union, Lancaster University, BBC, Markenfilm, VTT Technical Research Center and Pioneer Digital Design Center Limited.

Video distribution

Even if several of the existing P2P applications do support video distribution, rarely have they had a consistent infrastructure and the main goal to permit subscribing to video content. Several do provide searches but a careful integration of a video player/media manager with the distribution is so far extremely rare and an evolving proposition.

Considering the door the centralized systems like YouTube, Google Video and their ilk opened in regards to online video, this new need for distribution has lead to the creation of several P2P offering, like Joost (from the same team as the Kazaa and Skype applications), Tape it off the Internet or Veoh TV.

Of particular notice among the new emerging platforms is Miro (formerly Democracy Player), one of the first projects of this kind. Available at http://www.getmiro.com/ , it is open-source (supports all current versions of Windows, Mac OS and several Linux distro) non-profit video player that supports practically every format. Created by the Participatory Culture Foundation (PCF, a non-profit organization), it automates subscribing to video RSS feeds that are then downloaded using Bittorrent that reduces the money spent on bandwidth required for distribution. Available is also the Miro Video Converter application, that promotes the video conversion to the patent-free OGG Theora codec (which provides the same quality of the non free h.264) and has been backing the campaign to get Video on Wikipedia.

Real time video

Transmission of live events to millions of people using the actual infrastructure imposes limits on the quality of the output and high expectations on the hardware resources, not only on network resources but on the encoding and playback capabilities on each side of the transfer.

This has lead to the use of P2P network, as an attempt to save server bandwidth. An example of this new approach is the MSR Asia Peer-to-peer Video Broadcast System ( http://research.microsoft.com/en-us/projects/p2pbroadcast/ ) from Microsoft Research Asia that was used in the 2008 Beijing Olympic Games claiming the use of more than two million Internet peers.

Distribute digital content over semi-private P2P networks

P2P has also been used in recent times to help ease the load on distribution digital content in general in a system of restricted participation (non-public access) and dedicated function.

BitTorrent DNA or BitTorrent Delivery Network Accelerator, reportedly available since January 2005 operates in this fashion. This distribution technology has been especially interesting for games software houses, that have been moving their business to the Internet and adopting the business practice of subscription to services or content, in place of selling the ownership of a copy of the digital goods they create, or in a mix model. This will require a network connection to download extra content often protected by DRM, the number of transfers increases, in a way that it is not uncommon that during the update process or even the installation of the product the user's machine will be utilized to create a P2P network, sometimes even without the full knowledge of the consumer, to help the distribution of that paid content.

Blizzard for instance, uses BitTorrent and does inform users that when downloading some contents the user will also be sharing it across a P2P network.

Solid State Networks already offers a P2P-based delivery solution as does Akamai, that sells Netsession Interface to game publishers. There are also other tools that are used to the same effect (not exclusive to games) but as helpers to the distribution of large files, Pando Media Booster is an example.

There is however a problems about transparency over the utilization of the costumers' resources. Solid State Networks recently, and after several costumers reported issues regarding this new approach, that often degrades users systems security, stability and performance. Has started a campaign to establish an industry "best practices", or most of this solutions will fall in the definition of malware, http://www.solidstatenetworks.com/index.php/about-us/p2p-best-practices/ (PDF).

P2P is also used for digital art, for instance the Electric sheep project ( http://community.electricsheep.org ), a C++ open source based on libtorrent BitTorrent implementation, is a distributed computing project for animating and evolving fractal flames, which are in turn distributed to the networked computers, which display them as a screensaver. The process is transparent to the casual user, who can simply install the software as a screensaver. Alternatively, the user may become more involved with the project, manually creating sheep (video files of animated fractal flames) for upload to the server.

Hardware

As P2P protocols become more mature and claims to have a good level of public adoption, that we will see more embedded devices supporting the P2P solutions. This is already beginning to happen:

• TonidoPlug ( http://www.tonidoplug.com ) a tiny, low-power, low-cost personal home server and NAS device. Tonido Torrent is a web-based torrent client based on the BitTorrent implementation in the Boost licensed libtorrent library.
• Excito ( http://excito.com ) offers also a range of servers, named "Buba" that supports the BitTorrent protocol by using the Boost licensed libtorrent library.

Set Top Boxes

P2P technologies can also be used to provide a means to at low cost distribute content in an automated way.

Using a peer to peer architecture directly connected to a broadband line, a set top box (a stripped down PC of sorts), with an operating software and some storage space can for instance provide a service similar to video on demand.

VUDU

VUDU ( http://www.vudulabs.com/ ), thousands of movies delivered directly to your TV, it doesn't require a PC and is independent of your cable or satellite TV service.

Distributed File-systems

Distributed File-systems aren't new but pre-P2P system depended on a server (or the election of a server from a pool of known machines) and were prominently focused on LANs that provided increased stability to the network. New systems are more reliable facing the volatility of a network and implement the new technologies P2P relies on. Most implementation of a P2P distributed File-systems will have evolved based on the FreeNet model to some degree.

Tahoe Least-Authority Filesystem (Tahoe-LAFS)

Tahoe-LAFS (http://allmydata.org/trac/tahoe/wiki) a secure remote (distributed) filesystem, released under the GNU's General Public License (GPL), that shares with P2P the underlying network architecture and the principle of least authority, but it is not entirely decentralized. It requires a central node, called an Introducer, needed to connect new nodes.

With the objective of creating a fault-tolerant storage pool across several peers (cloud storage) were everybody provides storage for each other. The files are distributed across the multiple nodes using AES encryption. A variation of Reed-Solomon error correction is used to permit peers to disconnect without affecting the integrity of the content.

Tahoe is free software: all of the source code is available under an open-source license. The home page is at http://allmydata.org . Tahoe is sponsored by allmydata.com, which uses it as the back-end for a commercial personal-data backup service.

Omemo (http://www.omemo.com/)

Omemo (http://www.omemo.com/) is a free and open source (Visual Basic) P2P application under the GPL, from Pablo Soto creator of the MANOLITO protocol and Blubster. Omemo takes a different approach and uses a ring-shaped DHT based on Chord. It is meant to support key based routing while keeping query source obscurity due to randomization. Available for Windows.

Mobile Peer-to-Peer Computing

PEPERS Project

The PEPERS project (http://www.pepers.org/) focus is to design, implement and validate a reliable platform with high-level support for the design, development and operational deployment of secure mobile peer-to-peer applications for future Ambient Intelligent (AmI) environments. The platform will greatly assist the work and benefit both mobile application and service providers and service users. The project will address issues related to security, privacy, trust and access control in mobile peer-to-peer (p2p) systems by proposing a relevant framework architecture. The framework will include support for policy-based security management for mobile systems. The specific thematic focus of the project (that reflects also in the selection of the pilot scenarios that will be implemented) is the collaboration among teams dispersed over a geographical area. The consortium partners come from 4 EU Member states (Greece, UK, Italy and Cyprus) and include key technology providers and industry players, and leading academic institutions, as well as users partners from diverse business domains (media and journalism, security services) with increased needs for secure collaboration through advanced technological solutions that bring significant expertise and knowledge in PEPERS related technologies as well as the underlying operating business models.

From a technical point of view the project will focus on:

• the definition of appropriate security services for mobile peer-to-peer applications over suitable protocols
• the analysis and design of possible platforms and interfaces in mobile devices that can provide security services to support peer-to-peer applications
• the definition of interfaces based on open standards that will allow secure mobile access to application servers

The secure use of mobile peer-to-peer applications based on the project technology will be validated in real-life pilot applications supporting collaborative work in two operational domains:

• media and journalism: reporters working on assignment need to be able to record, edit and exchange information in a way that will protect and monitor intellectual property rights and the interests of the organization that employs them. This is particularly relevant as it covers both general issues of a mobile workforce and more specific issues related to the media application domain.
• physical security: guards and mobile patrols need to be able to receive and transmit sensitive customer information in a dynamic environment when they are called to respond and co-operate in case of ad-hoc exceptional situations.

Digital Currency

Bit Coin

Bitcoin ( http://www.bitcoin.org ), a peer-to-peer network based digital currency system without a central server or trusted parties, no central authority to issue new money or keep track of transactions. Nodes of the network hold the crypto keys to their own money and transact directly with each other, with the help of the network to check for double-spending.

The limited inflation of the Bitcoin system’s money supply is distributed evenly (by CPU power) throughout the network.

• Money transfers through the Internet, without middlemen.
• No third parties control over transactions.
• Transactions are practically free.
• No inherent instability caused by fractional reserve banking or policies of central banks.