A key-order preserving structured overlay network is a class of structured overlay network that preserves, in its structure, the order of keys to support efficient range queries. This paper presents a novel key-order preserving structured overlay network “Suzaku”. Similar to the conventional Chord#, Suzaku uses a periodically updated finger table as a routing table, but extends its uni-directional finger table to bi-directional, which achieves ⌈log2 n⌉-1 maximum lookup hops in the converged state. Suzaku introduces active and passive bi-directional finger table update algorithms for node insertion and deletion. This method maintains good lookup performance (lookup hops increase nearly logarithmically against n) even in churn situations. As well as its good performance, the algorithms of Suzaku are simple and easy to implement. This paper describes the principles of Suzaku, followed by simulation evaluations, in which it showed better performance than the conventional networks, Chord# and Skip Graph.

With the rapid rise in the demand for location related service, communication devices such as PDAs or cellar phones must be able to search and manage information related to the geographical location. To leverage location-related information is useful to get an in-depth perspective on environmental circumstances, such as traffic conditions or weather information. To handle the large number of information and queries communication devices generate in the current ubiquitous environment, some scalable mechanism must be required. DHTs and some overlay networks supporting range search are proposed. However, these overlay networks can not process queries of geographical region search. In this paper, we propose a overlay network called "Mill" which can efficiently manage information related to the geographical location. In DHT based overlay networks, each node has responsibility to manage a part of the whole hash table. DHTs provide scalable systems and support fast search. However, DHTs are not good at solving geographical search (range search), because hash function only supports exact match. In the Mill network, each node manages a part of ID-space calculated by "Z-ordering," which represents squire surface of the earth. This structure of ID-space enables to process region queries easily and fast. And Mill supports any scale of region search. We evaluate proposed system by using traffic infomation generator called "HAKONIWA." Simulation results show that the performance of Mill is good as well as other DHT systems. In addtion, Mill provedes more efficient region search than other overlay networks supporting range search.

There is an increasing requirement for supporting complex multidimensional queries in Peer-to-Peer systems. In the centralized spatial database world, R-Trees and its variant structures are widely accepted due to their capabilities to manage complex multidimensional queries. In this paper, we propose a new multidimensional indexing structure for P2P systems, called Distributed Hilbert R-Trees (DHR-Trees), in which peers organize themselves into an overlay network, dynamically maintain routing tables with region information and collaboratively execute complex multidimensional queries, such as range query and k-nearest neighbors query, efficiently. DHR-Trees has similar topology to the P-Trees P2P system. The peers' routing tables are enhanced with spatial region information, which allow multidimensional query predicates to be adapted into P2P systems with minor modification. The structure design and two major multidimensional query algorithms are presented. Our experimental results demonstrate that it performs well on range queries and k-nearest neighbors queries with multidimensional data set.

File-sharing Peer-to-Peer systems are effective for autonomous data retrieval and provision over the networks. However, the early data retrieval schemes such as Gnutella and Local Indices have low performance and large overhead. In order to solve weakness of early schemes, this paper proposes a dynamic scheme for data retrieval and provision, in which indices are adaptively allocated in appropriate nodes to variation of traffic patterns caused by query messages. The simulation experimental results show that the proposed scheme has good performance with reasonable overhead even when the traffic patterns vary as time proceeds.

In this paper, we propose a new method to enhance the fault-tolerance of the Content Addressable Network (CAN), which is known as a typical pure P2P system based on the notion of Distributed Hash Table (DHT). The basic idea of the proposed method is to introduce redundancy to the management of index information distributed over the nodes in the given P2P network, by allowing each index to be assigned to several nodes, which was restricted to be one in the original CAN system. To keep the consistency among several copies of indices, we propose an efficient synchronization scheme based on the notion of labels assigned to each copy in a distinct manner. The performance of the proposed scheme is evaluated by simulation. The result of simulations indicates that the proposed scheme significantly enhances the fault-tolerance of the CAN system.

With the rapid spread of information and ubiquitous access of browsers, flash crowds, a sudden, unanticipated surge in the volume of request rates, have become the bane of many Internet websites. This paper models and presents FCAN, an adaptive network that dynamically optimizes the system structure between peer-to-peer (P2P) and client-server (C/S) configurations to alleviate flash crowds effect. FCAN constructs P2P overlay on cache proxy server layer to distribute the flash traffic from origin server. It uses policy-configured DNS redirection to route the client requests in balance, and adopts strategy load detection to monitor and react the load changes. Our preliminary simulation results showed that the system is overall well behaved, which validates the correctness of our design.

Peer-to-peer (P2P) web caching has been studied recently as it can exploit local caches of peers for web caching without additional infrastructure. However, dynamic join/leave behaviors or local caching strategies of the peers due to their autonomy in a P2P network may limit the performance of P2P web caching. To overcome these limitations, we propose an effective directory-based P2P web caching system under dynamic participation of peers. We introduce the object lifetime in a P2P network considering the lifetimes of both an object in the local cache of a peer and a peer who owns the object, and utilize this object lifetime information for neighbor selection and storage management in the P2P web caching framework. For the neighbor selection, the proposed system utilizes the object lifetimes in selecting accurately a neighbor who would still retain the requested object and still remain in the P2P network. To improve the storage management, the proposed system uses efficiently the entire cache storage of the P2P network in such a way that the object is stored selectively in the local cache of the peer who requested it, considering the object lifetime. The trace-driven simulation results show that the proposed system has higher accuracy and fewer redirection failures than the conventional directory-based P2P web caching system in the feasible P2P network.

A framework is described for a peer-to-peer information exchange system, and a collaborative information retrieval (IR) scheme for the system is proposed. The aims of the system include smooth knowledge and information management to activate organizations or communities. Conventional server-centric systems are weak because they create information-providing bottlenecks. Accordingly, the proposed framework targets the collaborative inter-working of personal repositories that accumulate per-user information, and accept and service requests. Issues concerning the framework are addressed. One issue is the retrieval of information from another's personal repository; the retrieval criteria of a system are tightly personalized for its user. The system is assumed to employ a vector space model with a concept-base as its IR mechanism. The vector space on one system is very different from that on another system. Another issue is the automated control of the information-providing criteria. This paper presents solutions to the first problem. To achieve IR that provides satisfactory results to a user requiring information from another's personal repository, we need vector space equalization to compensate for the differences in the vector spaces of the personal repositories. The paper presents a vector space equalization scheme, the automated relevance feedback scheme, that compensates the differences in the vector spaces of the personal repositories. We implement the scheme as a system and evaluate its performance using documents on the Internet.