Tuning Vivaldi: Achieving Increased Accuracy and Stability

Elser, Benedikt; Forschler, A.; Fuhrmann, Thomas

doi:10.1007/978-3-642-10865-5_15

Cited by 3 publications

(5 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In [3] we showed that one can achieve very good results with this algorithm, depending on the chosen parameters. However, both algorithms are far from producing optimal results.…”

Section: The Vivaldi Algorithmmentioning

confidence: 93%

See 1 more Smart Citation

Spring for Vivaldi -- Orchestrating Hierarchical Network Coordinates

Elser¹,

Forschler

Fuhrmann³

2010

2010 IEEE Tenth International Conference on Peer-to-Peer Computing (P2P)

View full text Add to dashboard Cite

Network coordinates have become an important building block for efficient peer-to-peer applications. They allow the peers to determine those other peers to whom they are likely to experience a low round trip time. So they can preferably attach to these selected peers and thereby create efficient overlay structures. Despite the substantial interest in this topic, there is still room for improvement.In this paper, we present Hierarchical Vivaldi, a variant of the popular Vivaldi algorithm. It predicts the peers' embedding error so that it can optimize the peer selection process. We conducted extensive studies with several sets of Internet traffic traces. They compare our proposal to Vivaldi and its previously proposed optimization Pyxida. Our results show that Hierarchical Vivaldi can improve the accuracy of the RTT predictions by an order of magnitude.

show abstract

“…In [3] we showed that one can achieve very good results with this algorithm, depending on the chosen parameters. However, both algorithms are far from producing optimal results.…”

Section: The Vivaldi Algorithmmentioning

confidence: 93%

“…Furthermore, two additional data sets were used in our complete study in [7]. We used optimal parameters for both algorithms, derived from previous work [3]. We used a 4D space for R 0 , and a 2D space for R i with i > 0.…”

Section: Data Sets and Simulatormentioning

confidence: 99%

Spring for Vivaldi -- Orchestrating Hierarchical Network Coordinates

Elser¹,

Forschler

Fuhrmann³

2010

2010 IEEE Tenth International Conference on Peer-to-Peer Computing (P2P)

View full text Add to dashboard Cite

show abstract

“…First, we analyzed the structural correctness of the overlay. We used input data from the RocketFuel [37] project to create a realistic copy of the Nacamar/Tiscali provider network. Furthermore, the same code ran on PlanetLab using 118 nodes.…”

Section: Simulationmentioning

confidence: 99%

A generic KBR library with built-in simulation capabilities

Elser

Fuhrmann

2011

2011 International Conference on High Performance Computing &Amp; Simulation

Self Cite

View full text Add to dashboard Cite

Peer-to-peer technologies play an important role in modern applications. Structured P2P overlays have mostly been applied for distributed hash tables (DHTs). But structured routing overlays can provide more general services than a DHT alone. This fact is known in the literature, but rarely reflected in running code.In this paper, we introduce a generic key based routing (KBR) system. It provides a library with an easy-to-use interface and an underlying KBR daemon. Together, they can serve a wide variety of applications. The IGOR daemon and the library are implemented in a well-structured and extensible manner. An integrated network simulation component enables application developers to test their code locally before deployment. Our library has already been successfully used for various fully decentralized applications including a chat service [1], a video-on-demand system [2, 3], and a fully decentralized file system [4,5].

show abstract

“…This is done by getCandidates (Line 4) which only selects valid nodes of the two-hop neighborhood 7 of the given v base -node which is the starting point for each anchor search iteration. Following Song's approach, we choose a node minimizing the 7 Two hops are needed because of the alternating directions on the anchor tree relative prediction error e p (Lines 5-11), as defined in Equation 5. When no anchor can be found for a given base node because of non-available measurements, we choose another base node as a fallback.…”

Section: Fig 3: Embedding Visualizationmentioning

confidence: 99%

“…Vivaldi's performance has been further improved by Elser at al. [7]. Common to these approaches is that the accuracy is highly susceptible to triangle inequality violations (TIVs).…”

Section: Related Workmentioning

confidence: 99%

Bandwidth Prediction in the Face of Asymmetry

Schober

Brenner

Kapitza

et al. 2013

Distributed Applications and Interoperable Systems

View full text Add to dashboard Cite

Abstract. An increasing number of networked applications, like video conference and video-on-demand, benefit from knowledge about Internet path measures like available bandwidth. Server selection and placement of infrastructure nodes based on accurate information about network conditions help to improve the quality-of-service of these systems. Acquiring this knowledge usually requires fully-meshed ad-hoc measurements. These, however, introduce a large overhead and a possible delay in communication establishment. Thus, prediction-based approaches like Sequoia have been proposed, which treat path properties as a semimetric and embed them onto trees, leveraging labelling schemes to predict distances between hosts not measured before. In this paper, we identify asymmetry as a cause of serious distortion in these systems causing inaccurate prediction. We study the impact of asymmetric network conditions on the accuracy of existing tree-embedding approaches, and present direction-aware embedding, a novel scheme that separates upstream from downstream properties of hosts and significantly improves the prediction accuracy for highly asymmetric datasets. This is achieved by embedding nodes for each direction separately and constraining the distance calculation to inversely labelled nodes. We evaluate the effectiveness and trade-offs of our approach using synthetic as well as real-world datasets.

show abstract

Tuning Vivaldi: Achieving Increased Accuracy and Stability

Cited by 3 publications

References 6 publications

Spring for Vivaldi -- Orchestrating Hierarchical Network Coordinates

Spring for Vivaldi -- Orchestrating Hierarchical Network Coordinates

A generic KBR library with built-in simulation capabilities

Bandwidth Prediction in the Face of Asymmetry

Contact Info

Product

Resources

About