As broadband Internet speeds continue to increase, the home wireless ("WiFi") network may more frequently become a performance bottleneck. Past research, now nearly a decade old, initially documented this phenomenon through indirect inference techniques, noting the prevalence of WiFi bottlenecks but never directly measuring them. In the intervening years, access network (and WiFi) speeds have increased, warranting a re-appraisal of this important question, particularly with renewed private and federal investment in access network infrastructure. This paper studies this question, developing a new system and measurement technique to perform direct measurements of WiFi and access network performance, ultimately collecting and analyzing a first-of-its-kind dataset of more than 13,000 joint measurements of WiFi and access network throughputs, in a real-world deployment spanning more than 50 homes, for nearly two years. Using this dataset, we re-examine the question of whether, when, and to what extent a user's home wireless network may be a performance bottleneck, particularly relative to their access connection. We do so by directly and continuously measuring the user's Internet performance along two separate components of the Internet path -- from a wireless client inside the home network to the wired point of access (e.g., the cable modem), and from the wired point of access to the user's ISP. Confirming and revising results from more than a decade ago, we find that a user's home wireless network is often the throughput bottleneck. In particular, for users with access links that exceed 800~Mbps, the user's home wireless network was the performance bottleneck 100% of the time.
相關內容
Networking：IFIP International Conferences on Networking。
Explanation：國際網絡(luo)會議。
Publisher：IFIP。
SIT：
Blockchains are decentralized systems that provide trustable execution guarantees. Smart contracts are programs written in specialized programming languages running on blockchains that govern how tokens and cryptocurrency are sent and received. Smart contracts can invoke other smart contracts during the execution of transactions always initiated by external users. Once deployed, smart contracts cannot be modified, so techniques like runtime verification are very appealing for improving their reliability. However, the conventional model of computation of smart contracts is transactional: once operations commit, their effects are permanent and cannot be undone. In this paper, we proposed the concept of future monitors which allows monitors to remain waiting for future transactions to occur before committing or aborting. This is inspired by optimistic rollups, which are modern blockchain implementations that increase efficiency (and reduce cost) by delaying transaction effects. We exploit this delay to propose a model of computation that allows (bounded) future monitors. We show our monitors correct respect of legacy transactions, how they implement future bounded monitors and how they guarantee progress. We illustrate the use of future bounded monitors to implement correctly multi-transaction flash loans.
Memory bandwidth is known to be a performance bottleneck for FPGA accelerators, especially when they deal with large multi-dimensional data-sets. A large body of work focuses on reducing of off-chip transfers, but few authors try to improve the efficiency of transfers. This paper addresses the later issue by proposing (i) a compiler-based approach to accelerator's data layout to maximize contiguous access to off-chip memory, and (ii) data packing and runtime compression techniques that take advantage of this layout to further improve memory performance. We show that our approach can decrease the I/O cycles up to $7\times$ compared to un-optimized memory accesses.
We propose a novel stereo-confidence that can be measured externally to various stereo-matching networks, offering an alternative input modality choice of the cost volume for learning-based approaches, especially in safety-critical systems. Grounded in the foundational concepts of disparity definition and the disparity plane sweep, the proposed stereo-confidence method is built upon the idea that any shift in a stereo-image pair should be updated in a corresponding amount shift in the disparity map. Based on this idea, the proposed stereo-confidence method can be summarized in three folds. 1) Using the disparity plane sweep, multiple disparity maps can be obtained and treated as a 3-D volume (predicted disparity volume), like the cost volume is constructed. 2) One of these disparity maps serves as an anchor, allowing us to define a desirable (or ideal) disparity profile at every spatial point. 3) By comparing the desirable and predicted disparity profiles, we can quantify the level of matching ambiguity between left and right images for confidence measurement. Extensive experimental results using various stereo-matching networks and datasets demonstrate that the proposed stereo-confidence method not only shows competitive performance on its own but also consistent performance improvements when it is used as an input modality for learning-based stereo-confidence methods.
We consider the message complexity of verifying whether a given subgraph of the communication network forms a tree with specific properties both in the KT-$\rho$ (nodes know their $\rho$-hop neighborhood, including node IDs) and the KT-$0$ (nodes do not have this knowledge) models. We develop a rather general framework that helps in establishing tight lower bounds for various tree verification problems. We also consider two different verification requirements: namely that every node detects in the case the input is incorrect, as well as the requirement that at least one node detects. The results are stronger than previous ones in the sense that we assume that each node knows the number $n$ of nodes in the graph (in some cases) or an $\alpha$ approximation of $n$ (in other cases). For spanning tree verification, we show that the message complexity inherently depends on the quality of the given approximation of $n$: We show a tight lower bound of $\Omega(n^2)$ for the case $\alpha \ge \sqrt{2}$ and a much better upper bound (i.e., $O(n \log n)$) when nodes are given a tighter approximation. On the other hand, our framework also yields an $\Omega(n^2)$ lower bound on the message complexity of verifying a minimum spanning tree (MST), which reveals a polynomial separation between ST verification and MST verification. This result holds for randomized algorithms with perfect knowledge of the network size, and even when just one node detects illegal inputs, thus improving over the work of Kor, Korman, and Peleg (2013). For verifying a $d$-approximate BFS tree, we show that the same lower bound holds even if nodes know $n$ exactly, however, the lower bound is sensitive to $d$, which is the stretch parameter.
Strategies for partially observable Markov decision processes (POMDP) typically require memory. One way to represent this memory is via automata. We present a method to learn an automaton representation of a strategy using a modification of the L*-algorithm. Compared to the tabular representation of a strategy, the resulting automaton is dramatically smaller and thus also more explainable. Moreover, in the learning process, our heuristics may even improve the strategy's performance. In contrast to approaches that synthesize an automaton directly from the POMDP thereby solving it, our approach is incomparably more scalable.
The logic of information flows (LIF) has recently been proposed as a general framework in the field of knowledge representation. In this framework, tasks of procedural nature can still be modeled in a declarative, logic-based fashion. In this paper, we focus on the task of query processing under limited access patterns, a well-studied problem in the database literature. We show that LIF is well-suited for modeling this task. Toward this goal, we introduce a variant of LIF called "forward" LIF (FLIF), in a first-order setting. FLIF takes a novel graph-navigational approach; it is an XPath-like language that nevertheless turns out to be equivalent to the "executable" fragment of first-order logic defined by Nash and Lud\"ascher. One can also classify the variables in FLIF expressions as inputs and outputs. Expressions where inputs and outputs are disjoint, referred to as io-disjoint FLIF expressions, allow a particularly transparent translation into algebraic query plans that respect the access limitations. Finally, we show that general FLIF expressions can always be put into io-disjoint form.
Graph neural networks (GNNs) have demonstrated a significant boost in prediction performance on graph data. At the same time, the predictions made by these models are often hard to interpret. In that regard, many efforts have been made to explain the prediction mechanisms of these models from perspectives such as GNNExplainer, XGNN and PGExplainer. Although such works present systematic frameworks to interpret GNNs, a holistic review for explainable GNNs is unavailable. In this survey, we present a comprehensive review of explainability techniques developed for GNNs. We focus on explainable graph neural networks and categorize them based on the use of explainable methods. We further provide the common performance metrics for GNNs explanations and point out several future research directions.
Residual networks (ResNets) have displayed impressive results in pattern recognition and, recently, have garnered considerable theoretical interest due to a perceived link with neural ordinary differential equations (neural ODEs). This link relies on the convergence of network weights to a smooth function as the number of layers increases. We investigate the properties of weights trained by stochastic gradient descent and their scaling with network depth through detailed numerical experiments. We observe the existence of scaling regimes markedly different from those assumed in neural ODE literature. Depending on certain features of the network architecture, such as the smoothness of the activation function, one may obtain an alternative ODE limit, a stochastic differential equation or neither of these. These findings cast doubts on the validity of the neural ODE model as an adequate asymptotic description of deep ResNets and point to an alternative class of differential equations as a better description of the deep network limit.
Generative adversarial networks (GANs) have been extensively studied in the past few years. Arguably their most significant impact has been in the area of computer vision where great advances have been made in challenges such as plausible image generation, image-to-image translation, facial attribute manipulation and similar domains. Despite the significant successes achieved to date, applying GANs to real-world problems still poses significant challenges, three of which we focus on here. These are: (1) the generation of high quality images, (2) diversity of image generation, and (3) stable training. Focusing on the degree to which popular GAN technologies have made progress against these challenges, we provide a detailed review of the state of the art in GAN-related research in the published scientific literature. We further structure this review through a convenient taxonomy we have adopted based on variations in GAN architectures and loss functions. While several reviews for GANs have been presented to date, none have considered the status of this field based on their progress towards addressing practical challenges relevant to computer vision. Accordingly, we review and critically discuss the most popular architecture-variant, and loss-variant GANs, for tackling these challenges. Our objective is to provide an overview as well as a critical analysis of the status of GAN research in terms of relevant progress towards important computer vision application requirements. As we do this we also discuss the most compelling applications in computer vision in which GANs have demonstrated considerable success along with some suggestions for future research directions. Code related to GAN-variants studied in this work is summarized on //github.com/sheqi/GAN_Review.
Deep neural networks (DNNs) are successful in many computer vision tasks. However, the most accurate DNNs require millions of parameters and operations, making them energy, computation and memory intensive. This impedes the deployment of large DNNs in low-power devices with limited compute resources. Recent research improves DNN models by reducing the memory requirement, energy consumption, and number of operations without significantly decreasing the accuracy. This paper surveys the progress of low-power deep learning and computer vision, specifically in regards to inference, and discusses the methods for compacting and accelerating DNN models. The techniques can be divided into four major categories: (1) parameter quantization and pruning, (2) compressed convolutional filters and matrix factorization, (3) network architecture search, and (4) knowledge distillation. We analyze the accuracy, advantages, disadvantages, and potential solutions to the problems with the techniques in each category. We also discuss new evaluation metrics as a guideline for future research.
小貼士
登錄享
注冊地址： 北京市海淀區羊坊店路18號2幢3層301-191