Entity

Time filter

Source Type


Shang S.,Princeton University | Cuff P.,Hong Kong University of Science and Technology | Hui P.,Telekom Innovation Laboratories | Kulkarni S.,Princeton University
Proceedings - IEEE INFOCOM | Year: 2013

We analyze a class of distributed quantized consensus algorithms for arbitrary networks. In the initial setting, each node in the network has an integer value. Nodes exchange their current estimate of the mean value in the network, and then update their estimate by communicating with their neighbors in a limited capacity channel in an asynchronous clock setting. Eventually, all nodes reach consensus with quantized precision. We start the analysis with a special case of a distributed binary voting algorithm, then proceed to the expected convergence time for the general quantized consensus algorithm proposed by Kashyap et al. We use the theory of electric networks, random walks, and couplings of Markov chains to derive an O(N3 log N) upper bound for the expected convergence time on an arbitrary graph of size N, improving on the state of art bound of O(N4 log N) for binary consensus and O(N 5) for quantized consensus algorithms. Our result is not dependent on the graph topology. Simulations are performed to validate the analysis. © 2013 IEEE. Source


Li Y.,Tsinghua National Laboratory for Information Sciences and Technology | Qian M.,Tsinghua National Laboratory for Information Sciences and Technology | Jin D.,Tsinghua National Laboratory for Information Sciences and Technology | Hui P.,Hong Kong University of Science and Technology | And 5 more authors.
IEEE Transactions on Mobile Computing | Year: 2014

To cope with explosive traffic demands on current cellular networks of limited capacity, Disruption Tolerant Networking (DTN) is used to offload traffic from cellular networks to high capacity and free device-to-device networks. Current DTN-based mobile data offloading models are based on simple and unrealistic network assumptions which do not take into account the heterogeneity of mobile data and mobile users. We establish a mathematical framework to study the problem of multiple-type mobile data offloading under realistic assumptions, where (i) mobile data are heterogeneous in terms of size and lifetime; (ii) mobile users have different data subscribing interests; and (iii) the storages of offloading helpers are limited. We formulate the objective of achieving maximum mobile data offloading as a submodular function maximization problem with multiple linear constraints of limited storage, and propose three algorithms, suitable for the generic and more specific offloading scenarios, respectively, to solve this challenging optimization problem. We show that the designed algorithms effectively offload data to the DTN by using both the theoretical analysis and simulation investigations which employ both real human and vehicular mobility traces. © 2002-2012 IEEE. Source


Li Y.,Tsinghua National Laboratory for Information Sciences and Technology | Jin D.,Tsinghua National Laboratory for Information Sciences and Technology | Hui P.,Hong Kong University of Science and Technology | Hui P.,Telekom Innovation Laboratories | And 2 more authors.
IEEE Journal on Selected Areas in Communications | Year: 2016

With the emergence of demands for local area services, device-to-device (D2D) communication is proposed as a vital technology component for the next generation cellular communication to increase spectral efficiency and enhance system capacity. In a D2D underlaying cellular system, through carefully coordinated dynamic scheduling, some base stations can be shut down and the corresponding load can be transferred to the D2D communication, which provides significant energy savings without much impact on the overall system performance. In this paper, we study the base station scheduling in D2D communication given a large scale of users. By formulating this problem into a flow maximization problem that optimizes the data transmission from the base stations to the users, we obtain the optimal base station scheduling solution. Through extensive simulations with real human mobility traces, we demonstrate the effectiveness of our proposed scheme, which significantly enhances the system throughput compared with existing strategies. © 2015 IEEE. Source


Thakur G.S.,Oak Ridge National Laboratory | Hui P.,Telekom Innovation Laboratories | Helmy A.,University of Florida
GIS: Proceedings of the ACM International Symposium on Advances in Geographic Information Systems | Year: 2015

Transportation simulation technologies should accurately model traffic demand, distribution, and assignment parameters for urban environment simulation. These three parameters significantly impact transportation engineering benchmark process, are also critical in realizing realistic traffic modeling situations. In this paper, we model and characterize traffic density distribution of thousands of locations, intersection, and roadways around the world. The traffic densities are generated from millions of images collected over several years and processed using computer vision techniques. The resulting traffic density distribution time series are then analyzed. It is found using the goodness-of-fit test that the traffic density distributions follow heavy-tail models such as Weibull in over 90% of analyzed locations. Moreover, a heavy-tail gives rise to long-range dependence and self-similarity, which we studied by estimating the Hurst exponent (H). Our analysis based on seven different Hurst estimators strongly indicates that the traffic distribution patterns are stochastically self-similar (0.5 ≤ H ≤ 1.0). We believe this is an important finding that will influence the design and development of the next generation traffic simulation techniques and also aid in accurately modeling traffic engineering of urban systems. In addition, it shall provide a much-needed input for the development of smart cities. © 2015 ACM. Source


Uzun A.,TU Berlin | Salem M.,Telekom Innovation Laboratories | Kupper A.,TU Berlin
Proceedings - 2013 IEEE 7th International Conference on Semantic Computing, ICSC 2013 | Year: 2013

Conventional positioning approaches for Location-based Services (LBS) such as those provided by Google and Apple, are solely driven by geometric spatial data. Especially in proactive LBS scenarios, in which users are notified as soon as they reach a certain area, locations are mostly defined by geofences and do not incorporate any further information from the semantics of the location, such as the points of interest in the vicinity or more detailed information about the district the user is in. Leveraging LBS with the extensive pool of interconnected data in the Linking Open Data (LOD) Cloud will improve the LBS experience and will enable the development of sophisticated proactive services. In this paper, we present a Semantic Positioning Platform that enhances classic positioning methods by semantic features. This platform utilizes the OpenMobileNetwork, which is a Live Crowd sourcing Platform providing static as well as dynamic mobile network topology data based on the principles of Linked Data. It further uses the Positioning Enabler that enables persistent user background tracking and subscription to Semantic LBS Services. The Semantic Positioning approach allows LBS providers to locate users with respect to the semantics of their position instead of defining spatial geofences. As a proof-of-concept, a Restaurant Recommender Service is presented and its applicability is evaluated. © 2013 IEEE. Source

Discover hidden collaborations