New method improves search functionality in P2P networks

Peer-to-Peer (P2P) programs such as Kazaa and eDonkey have become extremely popular over a short period of time. As the amount of data in these networks is growing at a fast rate, the limited functionality of P2P programs often fails to meet the needs of the users. By integrating the technology of database systems in P2P systems, the search functionality of these applications can be drastically improved.

Publication date: 08-07-2010

 

Peer-to-Peer (P2P) programs such as Kazaa and eDonkey have become extremely popular over a short period of time. As the amount of data in these networks is growing at a fast rate, the limited functionality of P2P programs often fails to meet the needs of the users. By integrating the technology of database systems in P2P systems, the search functionality of these applications can be drastically improved. Ying Zhang of the Centrum Wiskunde & Informatica in Amsterdam developed a method to achieve this, which she described in her PhD thesis Efficient Distributed Query Processing on Heterogeneous XQuery Engines. Her method offers many advantages to users and is much less demanding on the network than the present systems. Zhang received her PhD degree on 8 July at the University of Amsterdam.

P2P networks are well adapted to the dynamism and volatility of the underlying computer networks, but they use only simple and limited search techniques. Among a huge number of files the user has to filter the right ones himself. The development of applications for automatic filtering is complicated because, among others, the data are distributed over a large number of different sources. Zhang developed the XQuery Remote Procedure Call (XRPC) method in which she combines the advantages of P2P networks and those of database systems. XRPC is applied to XML data, which is the most common format for exchanging data on the Internet.

The method of Zhang connects a large number of database systems with each other to carry out complex search queries jointly in an efficient way. Zhang concentrated her research on the complex problem having XML query languages work with XML data that are distributed over a large amount of computers. XRPC defines among others an extremely simple and scalable protocol so that different XML database system can communicate with each other efficiently over different P2P networks. The method developed by Zhang is implemented in the existing XML database systems MonetDB/XQuery. Alongside P2P networks, Zhang’s method can also be used on many other kinds of networks varying from traditional client server systems to the more exotic networks such as Data Cyclotron.