We will review some of the major results in random graphs and some of the more challenging open problems. We will cover algorithmic and structural questions. We will touch on newer models, including those related to the WWW.

Random graphs

We propose a natural scene statistic-based distortion-generic blind/no-reference (NR) image quality assessment (IQA) model that operates in the spatial domain. The new model, dubbed blind/referenceless image spatial quality evaluator (BRISQUE) does not compute distortion-specific features, such as ringing, blur, or blocking, but instead uses scene statistics of locally normalized luminance coefficients to quantify possible losses of “naturalness” in the image due to the presence of distortions, thereby leading to a holistic measure of quality. The underlying features used derive from the empirical distribution of locally normalized luminances and products of locally normalized luminances under a spatial natural scene statistic model. No transformation to another coordinate frame (DCT, wavelet, etc.) is required, distinguishing it from prior NR IQA approaches. Despite its simplicity, we are able to show that BRISQUE is statistically better than the full-reference peak signal-to-noise ratio and the structural similarity index, and is highly competitive with respect to all present-day distortion-generic NR IQA algorithms. BRISQUE has very low computational complexity, making it well suited for real time applications. BRISQUE features may be used for distortion-identification as well. To illustrate a new practical application of BRISQUE, we describe how a nonblind image denoising algorithm can be augmented with BRISQUE in order to perform blind image denoising. Results show that BRISQUE augmentation leads to performance improvements over state-of-the-art methods. A software release of BRISQUE is available online: http://live.ece.utexas.edu/research/quality/BRISQUE_release.zip for public use and evaluation.

/pdf/no-reference-image-quality-assessment-in-the-spatial-domain-2mnlvf25ig.pdf

No-Reference Image Quality Assessment in the Spatial Domain

An important aim of research on the blind image quality assessment (IQA) problem is to devise perceptual models that can predict the quality of distorted images with as little prior knowledge of the images or their distortions as possible. Current state-of-the-art “general purpose” no reference (NR) IQA algorithms require knowledge about anticipated distortions in the form of training examples and corresponding human opinion scores. However we have recently derived a blind IQA model that only makes use of measurable deviations from statistical regularities observed in natural images, without training on human-rated distorted images, and, indeed without any exposure to distorted images. Thus, it is “completely blind.” The new IQA model, which we call the Natural Image Quality Evaluator (NIQE) is based on the construction of a “quality aware” collection of statistical features based on a simple and successful space domain natural scene statistic (NSS) model. These features are derived from a corpus of natural, undistorted images. Experimental results show that the new index delivers performance comparable to top performing NR IQA models that require training on large databases of human opinions of distorted images. A software release is available at http://live.ece.utexas.edu/research/quality/niqe_release.zip.

/pdf/making-a-completely-blind-image-quality-analyzer-4rsald8tn3.pdf

Making a “Completely Blind” Image Quality Analyzer

Over the past decade, there has been an explosion of interest in network research across the physical and social sciences. For social scientists, the theory of networks has been a gold mine, yielding explanations for social phenomena in a wide variety of disciplines from psychology to economics. Here, we review the kinds of things that social scientists have tried to explain using social network analysis and provide a nutshell description of the basic assumptions, goals, and explanatory mechanisms prevalent in the field. We hope to contribute to a dialogue among researchers from across the physical and social sciences who share a common interest in understanding the antecedents and consequences of network phenomena.

/pdf/network-analysis-in-the-social-sciences-440fzedv4c.pdf

Network Analysis in the Social Sciences

I. the origin of species by means of natural selection

Federated learning (FL) is a distributed learning approach where a set of end-user devices participate in the learning process by acting on their isolated local data sets. Here, we process local data sets of users where worst-case optimization theory is used to reformulate the FL problem where the impact of local data sets in training phase is considered as an uncertain function bounded in a closed uncertainty region. This representation allows us to compare the performance of FL with its centralized counterpart, and to replace the uncertain function with a concept of protection functions leading to more tractable formulation. The latter supports applying a regularization factor in each user cost function in FL to reach a better performance. We evaluated our model using the MNIST data set versus the protection function parameters, e.g., regularization factors.

Representation of Federated Learning via Worst-Case Robust Optimization Theory

All-optical wavelength converters are central to next-generation exascale communication networks that span millions of processing cores. We examine the loss performance of a wavelength-routing optical packet switch by focussing on tunable wavelength converter parameters. Our cross-layer analysis reveals the stringent wavelength conversion requirements in terms of optical signal-to-noise ratio and conversion efficiency.

http://ieeexplore.ieee.org/iel7/6596164/6608232/06608277.pdf

Impact of tunable wavelength converter performance on all-optical wavelength-routing switches for data centers

This paper presents an analytical model for eval- uating the capability of wireless LANs (WLANs) to provision quantitative quality of service (QoS) guarantees. We consider a distributed medium access control (MAC) with class differentia- tion, where mobile nodes belonging to different classes may have heterogeneous traffic arrival processes or different contention windows. With on/off inputs, our analysis shows that the WLAN admission region under the QoS constraint can be significantly improved, when the statistical multiplexing effect is taken into account. Moreover, the statistical multiplexing gain can be further improved by aggregating the downlink flows at the access point (AP). We also demonstrate that the proper selection of contention windows plays an important role in improving the WLAN QoS capability, while the optimal contention window for each class and the maximum admission region can be jointly solved in our analytical model.

Improvement of WLAN QoS Capability via Statistical Multiplexing

The assessment of real-time intelligent transportation system (ITS) applications, such as traffic management and adaptive route guidance systems, requires the use of fast and near real-time dynamic traffic simulation models. Even off-line applications, used for testing planning scenarios, often require fast-enough traffic simulation models that enable the required repetitive simulations. This is even more critical for large-scale networks with millions of vehicles. This paper investigates the speedup of DTA simulation models, using compiler optimizations and parallelism. DynusT as a widely used DTA model was evaluated as a test case, while its results could be generalized because we have used real-networks and calibrated them using real data sets in the Greater Toronto and Hamilton Area (GTHA). Extensive testing is performed to evaluate various dimensions for speed-up including: network size, number of processors, various optimization levels and operating systems. The performance results show that compiler optimizations and parallelism allow to: 1) double the speed required for a 4-hour simulation after 12 iterations to reach equilibrium, and 2) bring down the initial simulation time (required for network loading) by 2.5 times, enabling the testing of various real-time ITS applications.

Speedup of DTA-Based Simulation of Large Metropolises for Quasi Real-Time ITS Applications

Ensuring quality of service (QoS) guarantees in service systems is a challenging task, particularly when the system is composed of more fine-grained services, such as service function chains. An important QoS metric in service systems is the end-to-end delay, which becomes even more important in delay-sensitive applications, where the jobs must be completed within a time deadline. Admission control is one way of providing end-to-end delay guarantee, where the controller accepts a job only if it has a high probability of meeting the deadline. In this paper, we propose a reinforcement learning-based admission controller that guarantees a probabilistic upper-bound on the end-to-end delay of the service system, while minimizes the probability of unnecessary rejections. Our controller only uses the queue length information of the network and requires no knowledge about the network topology or system parameters. Since long-term performance metrics are of great importance in service systems, we take an average-reward reinforcement learning approach, which is well suited to infinite horizon problems. Our evaluations verify that the proposed RL-based admission controller is capable of providing probabilistic bounds on the end-to-end delay of the network, without using system model information.

Alberto Leon-Garcia

Papers

Representation of Federated Learning via Worst-Case Robust Optimization Theory

Impact of tunable wavelength converter performance on all-optical wavelength-routing switches for data centers

Improvement of WLAN QoS Capability via Statistical Multiplexing

Speedup of DTA-Based Simulation of Large Metropolises for Quasi Real-Time ITS Applications

Reinforcement Learning-based Admission Control in Delay-sensitive Service Systems