Systems, methods and programming for routing and indexing globally addressable objects and associated business models

Abstract

Methods, apparatus, and programming recorded in machine readable memory are provided for the index, search and retrieval of objects on a global network. This inventive system embeds a distributed index in a routing layer to enable fast search. The method provides dynamic insertion, lookup, retrieval, and deletion of participating nodes, objects and associated metadata in a completely decentralized fashion. Nodes can dynamically join and leave the network. This infrastructure can be applied to content networks for publishing, searching, downloading, and streaming.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS [0001] This application is a continuation of U.S. patent application Ser. No. 10 / 246,793, filed Sep. 18, 2002 and entitled “Systems, Methods and Programming for Routing and Indexing Globally Addressable Objects and Associated Business Models”, which application claims priority to U.S. Provisional Patent Application Ser. No. 60 / 323,354, filed Sep. 18, 2001.FIELD OF THE INVENTION [0002] The present invention relates to systems, methods and programming for routing and indexing globally addressable objects and associated business models. BACKGROUND OF THE INVENTION [0003] The rapid adoption of Internet access, coupled with the continuing increase in power of computing hardware, has created numerous new opportunities in network services. Nevertheless, the state of the network is still very similar to that of the Internet eight years ago when the web was introduced heavily based on the client-server model. [0004] In the last eight years, the trends ...

AI Technical Summary

Benefits of technology

[0008] Skyris Networks, Inc., the intended assignee of this patent application, has developed new, more efficient algorithms for distributed indexing which meet the needs of distributed network services by being scalable and fault tolerant. In this paper, we propose a new routing and search scheme which we call “transparent indexing”, in which the search index is embedded in a highly scalable, fault tolerant, randomized, distributed routing layer. This routing and indexing layer serves as a distributed directory service, a platform on which to build network services which share these properties and are reliable, efficient, and load balanced.

[0009] Our queuing-time based simulations show that SKYRIS can run efficiently on a network with billions of servers. Furthermore, the SKYRIS network can handle any distribution of hot spots (small pieces of popular content), or suddenly popular content, with minor latency compromises, and can do so transparently.

[0012] EFFICIENT RETRIEVAL: Our goal is to make SKYRIS as fast as the Web, if not faster. When doing a lookup on the web of a new domain, one first checks the DNS system, which can make several hierarchical contacts before serving you the IP address. The aim has been to keep the SKYRIS system similarly within several hops, where a hop consists of a message passed from one peer machine to the next. For larger files, there are other methods, such as retrieving from multiple sources simultaneously, which increase system throughput.

Problems solved by technology

However, most of these PCs are never using their full potential.

The client-server model suffers from numerous problems.

Servers are expensive to maintain, requiring money for hardware, bandwidth, and operations costs.

Traffic on the Internet is unpredictable.

Similarly, centralized sites may suffer from Denial-of-Service (DoS) attacks, which is malicious traffic that can take down a site by similarly flooding the site's connection to the network.

Furthermore, some network services, particularly bandwidth-intensive ones such as providing video or audio on a large scale, are simply impossible in a centralized model since the bandwidth demands exceed the capacity of any single site.

However, Freenet is relatively slow and it is rather difficult to actually download any content off of Gnutella, since the protocol does not scale beyond a few thousand hosts, so the amount of accessible content on the network is limited.

Furthermore, these systems are designed for scalability, but not explicitly for reliability, and their reliability is untested under real-world conditions running on unreliable machines.

Images