Results 11  20
of
70
Diagnosis of capacity bottlenecks via passive monitoring in 3G networks: an empirical analysis
 COMPUTER NETWORKS
, 2007
"... In this work we address the problem of inferring the presence of a capacity bottleneck from passive measurements in a 3G network. The study is based on one month of packet traces collected in the UMTS core network of mobilkom austria AG & Co KG, the leading mobile telecommunications provider in Aust ..."
Abstract

Cited by 5 (2 self)
 Add to MetaCart
In this work we address the problem of inferring the presence of a capacity bottleneck from passive measurements in a 3G network. The study is based on one month of packet traces collected in the UMTS core network of mobilkom austria AG & Co KG, the leading mobile telecommunications provider in Austria, EU. During the measurement period a bottleneck link in the UMTS core network was revealed and removed, therefore the traces enable the accurate analysis and comparison of the traffic behavior in the two network conditions: with and without a capacity bottleneck. Two approaches to bottleneck detection are investigated. The first one is based on the signal analysis of the marginal rate distribution of the traffic aggregate along one day cycle. Since TCPcontrolled traffic dominates the overall traffic mix, the presence of a bottleneck strains the aggregate rate distribution and compresses it against the capacity limit during the peak hour. The second approach is based on the analysis of several TCP performance parameters, e.g. estimated frequency of
Efficient and Dynamic Routing Topology Inference From EndtoEnd Measurements
"... Inferring the routing topology and link performance from a node to a set of other nodes is an important component in network monitoring and application design. In this paper we propose a general framework for designing topology inference algorithms based on additive metrics. The framework can flexi ..."
Abstract

Cited by 5 (0 self)
 Add to MetaCart
Inferring the routing topology and link performance from a node to a set of other nodes is an important component in network monitoring and application design. In this paper we propose a general framework for designing topology inference algorithms based on additive metrics. The framework can flexibly fuse information from multiple measurements to achieve better estimation accuracy. We develop computationally efficient (polynomialtime) topology inference algorithms based on the framework. We prove that the probability of correct topology inference of our algorithms converges to one exponentially fast in the number of probing packets. In particular, for applications where nodes may join or leave frequently such as overlay network construction, applicationlayer multicast, peertopeer file sharing/streaming, we propose a novel sequential topology inference algorithm which significantly reduces the probing overhead and can efficiently handle node dynamics. We demonstrate the effectiveness of the proposed inference algorithms via Internet experiments.
Multiple source multiple destination topology inference using network coding
 in Proc . of IEEE Symposium of Network Coding (NetCod
, 2009
"... Abstract — In this paper, we combine network coding and tomographic techniques for topology inference. Our goal is to infer the topology of a network by sending probes between a given set of multiple sources and multiple receivers and by having intermediate nodes perform network coding operations. W ..."
Abstract

Cited by 5 (1 self)
 Add to MetaCart
Abstract — In this paper, we combine network coding and tomographic techniques for topology inference. Our goal is to infer the topology of a network by sending probes between a given set of multiple sources and multiple receivers and by having intermediate nodes perform network coding operations. We combine and extend two ideas that have been developed independently. On one hand, network coding introduces topologydependent correlation, which can then be exploited at the receivers to infer the topology [1]. On the other hand, it has been shown that a traditional (i.e., without network coding) multiple source, multiple receiver tomography problem can be decomposed into multiple two source, two receiver subproblems [2]. Our first contribution is to show that, when intermediate nodes perform network coding, topological information contained in network coded packets allows to accurately distinguish among all different 2by2 subnetwork components, which was not possible with traditional tomographic techniques. Our second contribution is to use this knowledge to merge the subnetworks and accurately reconstruct the general topology. Our approach is applicable to any general Internetlike topology, and is robust to the presence of delay variability and packet loss. I.
Temporal Delay Tomography
 In Proceedings of the IEEE INFOCOM Conference
, 2008
"... Abstract—Multicastbased network tomography enables inference of average loss rates and delay distributions of internal network links from endtoend measurements of multicast probes. Recent work showed that this method, based on correlating observations of multicast receivers, also supports the inf ..."
Abstract

Cited by 5 (0 self)
 Add to MetaCart
Abstract—Multicastbased network tomography enables inference of average loss rates and delay distributions of internal network links from endtoend measurements of multicast probes. Recent work showed that this method, based on correlating observations of multicast receivers, also supports the inference of temporal loss characteristics of network links. In this paper, we show that temporal characteristics can, in fact, be estimated even for link delay processes. Knowledge of temporal delay characteristics has applications for delay sensitive services such as VoIP as well as for characterizing the queueing behavior of bottleneck links. By assuming mutually independent, but arbitrary link delay processes, we develop estimators which can infer, in addition to delay distributions, the probabilities of arbitrary patterns of delay, means and full distributions of delayrun periods at chosen delay levels, for each link in the multicast tree. By applying the recently proposed principle of subtreepartitioning, the estimator is made scalable to multicast trees of large degree. Estimation error and convergence rates are evaluated using simulations. I.
Embracing statistical challenges in the information technology age
 Technometrics
"... www.stat.berkeley.edu/users/binyu) This article examines the role of statistics in the age of information technology (IT). It begins by examining the current state of IT and of the cyberinfrastructure initiative aimed at integrating the technologies into science, engineering, and education to conver ..."
Abstract

Cited by 5 (0 self)
 Add to MetaCart
www.stat.berkeley.edu/users/binyu) This article examines the role of statistics in the age of information technology (IT). It begins by examining the current state of IT and of the cyberinfrastructure initiative aimed at integrating the technologies into science, engineering, and education to convert massive amounts of data into useful information. Selected applications from science and text processing are introduced to provide concrete examples of massive data sets and the statistical challenges that they pose. The thriving field of machine learning is reviewed as an example of current achievements driven by computations and IT. Ongoing challenges that we face in the IT revolution are also highlighted. The paper concludes that for the healthy future of our field, computer technologies have to be integrated into statistics, and statistical thinking in turn must be integrated into computer technologies. 1.
Network Routing Topology Inference from EndtoEnd Measurements
"... Abstract—Inference of the routing topology and link performance from a node to a set of other nodes is an important component of network monitoring and application design. In this paper we propose a general framework for designing topology inference algorithms based on additive metrics. Our framewor ..."
Abstract

Cited by 4 (0 self)
 Add to MetaCart
Abstract—Inference of the routing topology and link performance from a node to a set of other nodes is an important component of network monitoring and application design. In this paper we propose a general framework for designing topology inference algorithms based on additive metrics. Our framework allows the integration of both endtoend packet probing measurements and traceroute type measurements. Based on this framework we design several computationally efficient topology inference algorithms. In particular, we propose a novel sequential topology inference algorithm to address the probing scalability problem and handle dynamic node joining and leaving. We provide sufficient conditions for the correctness of our algorithms and derive lower bounds on the probability of correct topology inference. We conduct Internet experiments to evaluate and demonstrate the effectiveness of our algorithms. I.
Topology Discovery of Sparse Random Graphs With Few Participants ∗
, 2011
"... We considerthe taskoftopologydiscoveryofsparserandomgraphsusing endtoendrandom measurements(e.g., delay)between a subset ofnodes, referredto as the participants. The rest of the nodes are hidden, and do not provide any information for topology discovery. We consider topology discovery under two ro ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
We considerthe taskoftopologydiscoveryofsparserandomgraphsusing endtoendrandom measurements(e.g., delay)between a subset ofnodes, referredto as the participants. The rest of the nodes are hidden, and do not provide any information for topology discovery. We consider topology discovery under two routing models: (a) the participants exchange messages along the shortest paths and obtain endtoend measurements, and (b) additionally, the participants exchange messages along the second shortest path. For scenario (a), our proposed algorithm results in a sublinear editdistance guarantee using a sublinear number of uniformly selected participants. For scenario (b), we obtain a much stronger result, and show that we can achieve consistent reconstruction when a sublinear number of uniformly selected nodes participate. This implies that accurate discovery of sparse random graphs is tractable using an extremely small number of participants. We finally obtain a lower bound on the number of participants required by any algorithm to reconstruct the original random graph up to a given edit distance. We also demonstrate that while consistent discovery is tractable for sparse random graphs using a small number of participants, in general, there are graphs which cannot be discovered by any algorithm even with a significant number of participants, and with the availability of endtoend information along all the paths between the participants.
Network coding tomography for network failures
 in Proc. of INFOCOM, miniconference
, 2010
"... Abstract—Network Tomography (or network monitoring) uses endtoend pathlevel measurements to characterize the network, such as topology estimation and failure detection. This work provides the first comprehensive study of passive network tomography in the presence of network failures under the set ..."
Abstract

Cited by 3 (3 self)
 Add to MetaCart
Abstract—Network Tomography (or network monitoring) uses endtoend pathlevel measurements to characterize the network, such as topology estimation and failure detection. This work provides the first comprehensive study of passive network tomography in the presence of network failures under the setting that all nodes perform random linear network coding. In particular, we show that it is both necessary and sufficient for all nodes in the network to share common randomness, i.e., all local coding coefficients are chosen using a commonly shared random codebook. Without such common randomness, we prove that in the presence of adversarial or random failures, it is either theoretically impossible or computationally intractable to accurately estimate the topology of general networks, and then locate the failures. With such common randomness, we present several sets of algorithms for topology estimation and failure detection, under various settings of adversarial and random failures. For some scenarios our algorithms have polynomial timecomplexity, while for others we demonstrate computational intractability. Our main observation from this work is that the linear transforms arising from random linear network coding have some very specific relationships with the network structure, and these relationships can be leveraged to significantly aid
Statistical Aspects of the Analysis of Data Networks
"... Assessing and monitoring the performance of computer and communications networks is an important problem for network engineers. There has been a considerable amount of work on tools and techniques for data collection, modeling, and analysis within the network research community. The goal of this pap ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
Assessing and monitoring the performance of computer and communications networks is an important problem for network engineers. There has been a considerable amount of work on tools and techniques for data collection, modeling, and analysis within the network research community. The goal of this paper is to present an overview of the engineering problems and statistical issues, describe recent research developments, and summarize ongoing work and areas for further research. While there are many interesting issues related to network analysis, our focus here is on estimating and monitoring network QualityofService parameters. We discuss methods for estimating edgelevel parameters from endtoend pathlevel measurements, an important engineering problem that raises interesting statistical modeling issues. Other topics include network monitoring, network visualization, and discovering network topology. Data from a corporate network are used to illustrate the problems and techniques. As in any overview paper, the discussion is likely to be slanted towards our own research interests.