Peer-to-peer live streaming systems allow a bandwidthconstrained source to broadcast a video feed to a large number of users. In addition, a design with high link utilization can achieve high stream rates, supporting highquality video. Until now, only tree-based designs have been shown to achieve close-to-optimal rates in real-life conditions, leaving the question open as to the attainable efficiency of completely unstructured mesh-based approaches.In this paper we answer that question by showing that a carefully-designed mesh-based system can achieve closeto-optimal stream rates. Specifically, we implement and evaluate a design based on a mesh-based algorithm called DP/LU. Contrary to tree-based designs, DP/LU uses an unstructured overlay, which is easier to construct and is highly resistant to churn. In addition, we introduce mechanisms for overlay rewiring and source scheduling that lead to significant performance improvements.Our experimental evaluation shows that our design achieves 95% of the maximum achievable stream rate in a static environment, and 90% under high churn. This demonstrates that mesh-based designs are an excellent choice for scalable and robust high-quality peer-to-peer live streaming.
We consider a content delivery architecture based on geographically dispersed groups of "last-mile" CDN servers, e.g., set-top boxes located within users' homes. These servers may belong to administratively separate domains, such as multiple ISPs. We propose a set of scalable, adaptive mechanisms to jointly manage content replication and request routing within this architecture. Relying on primal-dual methods and fluid-limit techniques, we formally prove the optimality of our design. We further evaluate its performance on both synthetic and trace-driven simulations, based on real BitTorrent traces, and observe a reduction of network costs by more than 50% over traditional mechanisms such as LRU/LFU with closest request routing.
No abstract
Vehicular ad-hoc networks present great opportunity for information exchange and equal opportunity for abuse. Validating traffic information without imposing significant communication overheads is a hard problem. In this paper, we propose a solution for validating aggregated data. The main idea is to use random checks to probabilistically catch the attacker, and thereby discourage attacks in the network. Our solution relies on PKI based authentication and assumes a tamper-proof service in each car to carry out certain secure operations such as signing and timestamping. We try to keep the set of secure operations as small as possible, in accordance with the principle of economy of mechanism. We show that our solution provides security without significant communication overheads.
Abstract. We introduce Pastis, a completely decentralized multi-user read-write peer-to-peer file system. In Pastis every file is described by a modifiable inode-like structure which contains the addresses of the immutable blocks in which the file contents are stored. All data are stored using the Past distributed hash table (DHT), which we have modified in order to reduce the number of network messages it generates, thus optimizing replica retrieval. Pastis' design is simple compared to other existing systems, as it does not require complex algorithms like Byzantine-fault tolerant (BFT) replication or a central administrative authority. It is also highly scalable in terms of the number of network nodes and users sharing a given file or portion of the file system. Furthermore, Pastis takes advantage of the fault tolerance and good locality properties of its underlying storage layer, the Past DHT. We have developed a prototype based on the FreePastry open-source implementation of the Past DHT. We have used this prototype to evaluate several characteristics of our file system design. Supporting the close-toopen consistency model, plus a variant of the read-your-writes model, our prototype shows that Pastis is between 1.4 to 1.8 times slower than NFS. In comparison, Ivy and Oceanstore are between two to three times slower than NFS.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.