The quality of service for latency dependent content, such as video streaming, largely depends on the distance and available bandwidth between the consumer and the content. Poor provision of these qualities results in reduced user experience and increased overhead. To alleviate this, many systems operate caching and replication, utilising dedicated resources to move the content closer to the consumer. Latency-dependent content creates particular issues for community networks, which often display the property of strong internal connectivity yet poor external connectivity. However, unlike traditional networks, communities often cannot deploy dedicated infrastructure for both monetary and practical reasons. To address these issues, this paper proposes Corelli, a peer-to-peer replication infrastructure designed for use in community networks. In Corelli, high capacity peers in communities autonomously build a distributed cache to dynamically pre-fetch content early on in its popularity lifecycle. By exploiting the natural proximity of peers in the community, users can gain extremely low latency access to content whilst reducing egress utilisation. Through simulation, it is shown that Corelli considerably increases accessibility and improves performance for latency dependent content. Further, Corelli is shown to offer adaptive and resilient mechanisms that ensure that it can respond to variations in churn, demand and popularity.

Many Peer-to-Peer information retrieval systems that use a global index have already been proposed that can retrieve documents relevant to a query. Since documents are allocated to peers regardless of the query, the system needs to connect many peers to gather the relevant documents. We propose a new data allocation scheme for P2P information retrieval that we call Concordia. Concordia uses a node to allocate a document based on the weight of each term in the document to efficiently assemble all the documents relevant to a query from the P2P Network. Moreover, the node encodes the binary data of a document with an erasure code, and Concordia produces an efficient redundancy for counteracting node failures.

Abstract—When applied in a commercial deployment, DHTbased P2P protocols face a dilemma: although most real-world participants are so unstable that the maintenance overhead is prohibitively high, they must be effectively utilized due to the lack of stable participants. Thus, determining how to leverage unstable nodes to enhance system scalability and then maximize stability in high-churn scenarios becomes a substantial problem. This paper focuses on this topic, and our main findings are two folds: 1) we propose a homogeneous grouping scheme for scalability enhancement. Besides extending system storage capacity by admitting all nodes, it clusters homogeneous nodes together, deploys the inter- and intra-group connections distinctively, and tunes the number of groups, which aims to facilitate search efficiency; 2) we further look into how to maximize stability under this scheme, which is formulated as the problem Maximum Stability of Grouping. It not only proves to be NP-hard, but also infeasible; therefore, we propose an approximated grouping approach and reduce it to an optimization problem that proves to be feasible. Simulation results exhibit that our grouping strategy effectively captures the stability-scalability tradeoff. Based on our proposed measurement metrics, it doubles the storage capacity of so-called GiantOnly strategy by incurring slightly more churn and search latency, and is about four times as stable as Chord with equal capacity and mild improvement in search efficiency. Index Terms—Peer-to-peer, distributed hash table, stability, scalability, grouping, homogeneity, high churn, optimization I.

Abstract. We have designed, implemented and evaluated a resource adaptive distributed information sharing system where automatic adjustments are made internally in our information sharing system in order to cope with varying resource consumption. CPU load is monitored and a light-weight trigger mechanism is used to avoid overload situations on a per-machine basis. Additional improvements are obtained by calculating what we call a utility score to better determine how the data structures in the system should be arranged. Our results show that resource adaptation is an efficient way of improving query throughput, and that it is most effective when the number of stored data items in the system is large or many queries are performed concurrently. By applying resource adaptation, we are able to significantly improve the performance of our information sharing system.