Results 1  10
of
31
Gossiping in Distributed Systems
"... Gossipbased algorithms were first introduced for reliably disseminating data in largescale distributed systems. However, their simplicity, robustness, and flexibility make them attractive for more than just pure data dissemination alone. In particular, gossiping has been applied to data aggregatio ..."
Abstract

Cited by 47 (0 self)
 Add to MetaCart
(Show Context)
Gossipbased algorithms were first introduced for reliably disseminating data in largescale distributed systems. However, their simplicity, robustness, and flexibility make them attractive for more than just pure data dissemination alone. In particular, gossiping has been applied to data aggregation, overlay maintenance, and resource allocation. Gossiping applications more or less fit the same framework, with often subtle differences in algorithmic details determining divergent emergent behavior. This divergence is often difficult to understand, as formal models have yet to be developed that can capture the full design space of gossiping solutions. In this paper, we present a brief introduction to the field of gossiping in distributed systems, by providing a simple framework and using that framework to describe solutions for various application domains.
Fast Information Spreading in Graphs with Large Weak Conductance
"... Gathering data from nodes in a network is at the heart of many distributed applications, most notably, while performing a global task. We consider information spreading among n nodes of a network, where each node v has a message m(v) which must be received by all other nodes. The time required for i ..."
Abstract

Cited by 17 (2 self)
 Add to MetaCart
(Show Context)
Gathering data from nodes in a network is at the heart of many distributed applications, most notably, while performing a global task. We consider information spreading among n nodes of a network, where each node v has a message m(v) which must be received by all other nodes. The time required for information spreading has been previously upperbounded with an inverse relationship to the conductance of the underlying communication graph. This implies high running times for graphs with small conductance. The main contribution of this paper is an information spreading algorithm which overcomes communication bottlenecks and thus achieves fast information spreading for a wide class of graphs, despite their small conductance. As a key tool in our study we use the recently defined concept of
Compositional gossip: a conceptual architecture for designing gossipbased applications
 ACM SIGOPS Operating Systems Review
, 2007
"... Most proposed gossipbased systems use an adhoc design. We observe a low degree of reutilization among this proposals. We present how this limits both the systematic development of gossipbased applications and the number of applications that can benefit from gossipbased construction. We posit tha ..."
Abstract

Cited by 14 (3 self)
 Add to MetaCart
(Show Context)
Most proposed gossipbased systems use an adhoc design. We observe a low degree of reutilization among this proposals. We present how this limits both the systematic development of gossipbased applications and the number of applications that can benefit from gossipbased construction. We posit that these reinventthewheel approaches poses a significant barrier to the spread and usability of gossip protocols. This paper advocates a conceptual design framework based upon aggregating basic and predefined building blocks (B 2). We show how to compose building blocks within our framework to construct more complex blocks to be used in gossipbased applications. The concept is further depicted with two gossipbased applications described using our building blocks.
Streaming in a Connected World: Querying and Tracking Distributed Data Streams
 SIGMOD'07
, 2007
"... Today, a majority of data is fundamentally distributed in nature. Data for almost any task is collected over a broad area, and streams in at a much greater rate than ever before. In particular, advances in sensor technology ..."
Abstract

Cited by 14 (5 self)
 Add to MetaCart
Today, a majority of data is fundamentally distributed in nature. Data for almost any task is collected over a broad area, and streams in at a much greater rate than ever before. In particular, advances in sensor technology
Complexity of Data Collection, Aggregation, and Selection for Wireless Sensor Networks
"... Processing the gathered information efficiently is a key functionality for wireless sensor networks. In this paper, we study the time complexity, message complexity, and energy cost complexity of various processing operations for a multihop wireless sensor network of n nodes. For most of the operat ..."
Abstract

Cited by 12 (6 self)
 Add to MetaCart
(Show Context)
Processing the gathered information efficiently is a key functionality for wireless sensor networks. In this paper, we study the time complexity, message complexity, and energy cost complexity of various processing operations for a multihop wireless sensor network of n nodes. For most of the operations studied in this paper, we first present a lowerbound on the complexity for the optimal methods, then we provide an (asymptotically matching) upperbound on the complexity by presenting efficient distributed algorithms to solve these problems. Let ϱT, ϱM, and ϱE be the approximation ratio of an algorithm in terms of time complexity, message complexity, and energy complexity respectively for a certain operation, such as data collection, data aggregation, or data selection. Specifically, we show that, for data collection, there are networks of n nodes and maximum degree ∆, such that ϱM ϱE = Ω(∆) for any algorithm. We then present an efficient algorithm for data collection with ϱT = O(1), ϱM = O(1), and ϱE = O(∆). For data aggregation, we show that there are networks of n nodes and maximum degree ∆, such that ϱT ϱE = Ω(∆) for any algorithm. We then present an efficient algorithm for data aggregation with ϱT = O(1), ϱM = O(1), and ϱE = O(∆). For data selection, we show that any deterministic distributed algorithm needs Ω( ∆ + D logD N) time to find the median of all data items, where N is the number of total elements collected by sensors. We then present a randomized algorithm that achieves this lowerbound with high probability. In terms of the message complexity, there is a graph G, such that Ω(n log h) messages are required to compute the kth smallest element in G in expectation and with probability at least 1/nδ for every constant δ < 1/2, where h = min(k, N − k). We also present a randomized algorithm that achieves this bound with high probability.
Whisper: Middleware for confidential communication in largescale networks
 in Distributed Computing Systems (ICDCS), 2011 31st International Conference on
, 2011
"... Abstract—A wide range of distributed applications requires some form of confidential communication between groups of users. In particular, the messages exchanged between the users and the identity of group members should not be visible to external observers. Classical approaches to confidential grou ..."
Abstract

Cited by 9 (2 self)
 Add to MetaCart
(Show Context)
Abstract—A wide range of distributed applications requires some form of confidential communication between groups of users. In particular, the messages exchanged between the users and the identity of group members should not be visible to external observers. Classical approaches to confidential group communication rely upon centralized servers, which limit scalability and represent single points of failure. In this paper, we present WHISPER, a fully decentralized middleware that supports confidential communications within groups of nodes in largescale systems. It builds upon a peer sampling service that takes into account network limitations such as NAT and firewalls. WHISPER implements confidentiality in two ways: it protects the content of messages exchanged between the members of a group, and it keeps the group memberships secret to external observers. Using multihops paths allows these guarantees to hold even if attackers can observe the link between two nodes, or be used as content relays for NAT bypassing. Evaluation in realworld settings indicates that the price of confidentiality remains reasonable in terms of network load and processing costs. I.
LOT: A Robust Overlay for Distributed Range Query Processing
"... Largescale datacentric services are often handled by clusters of computers that include hundreds of thousands of computing nodes. However, traditional distributed query processing techniques fail to handle the largescale distribution, peertopeer communication and frequent disconnection. In this ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
Largescale datacentric services are often handled by clusters of computers that include hundreds of thousands of computing nodes. However, traditional distributed query processing techniques fail to handle the largescale distribution, peertopeer communication and frequent disconnection. In this paper, we introduce LOT, a robust, faulttolerant and highly distributed overlay network for largescale peertopeer query processing. LOT is based on a robust tree overlay for distributed systems. It uses virtualization, replication, geographicbased clustering and flexible state definition as basic design principles. We show how we map these principles to desirable performance goals. Moreover, we provide a lightweight maintenance mechanism for updating state information. Analysis and simulations show that our approach is superior to other wellknown alternatives in its query processing performance and handling of churn. 1
1 Efficient Data Collection for Wireless Networks: Delay and Energy Tradeoffs
"... Abstract—In this paper, we study efficient data collection for wireless sensor networks. We present efficient distributed algorithms with approximately the minimum delay, or the minimum messages to be sent by all nodes, or the minimum total energy costs by all nodes. We analytically proved that all ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
(Show Context)
Abstract—In this paper, we study efficient data collection for wireless sensor networks. We present efficient distributed algorithms with approximately the minimum delay, or the minimum messages to be sent by all nodes, or the minimum total energy costs by all nodes. We analytically proved that all our methods are either optimum or are within constants factor of the optimum. We then investigate the possibility of designing one universal method such that the delay, the messages sent by nodes, and the total energy costs by all nodes are all optimum or within constants factor of optimum. Given a method A for data collection let T, M, and E be the approximation ratio of A in terms of time complexity, message complexity, and energy complexity respectively. We then show that, for data collection, there are networks of n nodes and maximum degree ∆, such that ME = Ω(∆) for any algorithm. Index Terms—Time complexity, message complexity, energy, sensor networks, data collection. I.
P.: Networkfriendly gossiping
 In: SSS ’09: Proceedings of the 11th International Symposium on Stabilization, Safety, and Security of Distributed Systems
, 2009
"... Abstract. The emergence of largescale distributed applications based on manytomany communication models, e.g., broadcast and decentralized group communication, has an important impact on the underlying layers, notably the Internet routing infrastructure. To make an effective use of network resou ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
Abstract. The emergence of largescale distributed applications based on manytomany communication models, e.g., broadcast and decentralized group communication, has an important impact on the underlying layers, notably the Internet routing infrastructure. To make an effective use of network resources, protocols should both limit the stress (amount of messages) on each infrastructure entity (routers, links), and balance as much as possible the load in the network. Most protocols use applicationlevel metrics such as delays to improve efficiency of content dissemination or routing, but the extend to which such applicationcentric optimizations help reduce and balance the load imposed to the infrastructure is unclear. In this paper, we elaborate on the design of such networkfriendly protocols and associated metrics. More specifically, we investigate randomnessbased gossip dissemination. We propose and evaluate different ways of making this representative protocol networkfriendly, while keeping its desirable properties (robustness and low delays). Simulations of the proposed methods using synthetic and real network topologies convey and compare their abilities to reduce and balance the load while keeping good performance. 1
Delay and Energy Efficiency Tradeoffs for Data Collections and Aggregation in Large Scale Wireless Sensor Networks
"... In this paper, we study efficient data collection for wireless sensor networks. We present efficient distributed algorithms with approximately the minimum delay, or the minimum messages to be sent by all nodes, or the minimum total energy consumed by all nodes. We analytically proved that all our m ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
In this paper, we study efficient data collection for wireless sensor networks. We present efficient distributed algorithms with approximately the minimum delay, or the minimum messages to be sent by all nodes, or the minimum total energy consumed by all nodes. We analytically proved that all our methods are either optimum or are within constants factor of the optimum. We then investigate the possibility of designing one universal method such that the delay, the messages sent by all nodes, and the total energy costs by all nodes are all optimum or are within constants factor of optimum. Given a method A for data collection, let ϱT, ϱM, and ϱE be the approximation ratio of A in terms of time complexity, message complexity, and energy complexity respectively. We then show that, for data collection, there are networks of n nodes and maximum degree ∆, such that ϱM ϱE = Ω(∆) for any algorithm. We also prove that our method achieves the best tradeoffs in terms of the time complexity, message complexity and the energy complexity.