The initial version of this paper was published in https://www.sciencedirect.com/science/article/abs/pii/S0920548906000122?via=ihub

In federated edge computing networks, edge computing (EC) nodes from multiple service providers (SPs) are installed to serve the same customer base and each SP intends to maximize their economic utilities. Therefore, to address research problems like EC node placement, job request allocation, and load balancing, we need to involve some economic aspects along with network engineering aspects. This implies that we need to propose novel system models and formulate problems in a significantly different way than the existing ones. Thus, we are motivated to discuss about the primary aspects of these new problem formulations and propose a novel economic and game-theoretic model for load balancing among federated EC nodes in this paper. In this game formulation, instead of focusing on latency minimization, under-loaded EC nodes intend to maximize their economic utilities by receiving any extra workload and incentives from their overloaded neighbors while satisfying the expected latency target. Furthermore, we design a centralized control mechanism, tailor-made for ultra-reliable and low latency (uRLL) applications, for implementing this load balancing framework by incorporating artificial intelligence (AI)-enhanced traffic prediction algorithms.

Background/foreground separation is one of the most fundamental tasks in computer vision, especially for video data. Robust PCA (RPCA) and its tensor extension, namely, Robust Tensor PCA (RTPCA), provide an effective framework for background/foreground separation by decomposing the data into low-rank and sparse components, which contain the background and the foreground (moving objects), respectively. However, in real-world applications, the video data is contaminated with noise. For example, in metal additive manufacturing (AM), the processed X-ray video to study melt pool dynamics is very noisy. RPCA and RTPCA are not able to separate the background, foreground, and noise simultaneously. As a result, the noise will contaminate the background or the foreground or both. There is a need to remove the noise from the background and foreground. To achieve the three terms decomposition, a smooth sparse Robust Tensor Decomposition (SS-RTD) model is proposed to decompose the data into static background, smooth foreground, and noise, respectively. Specifically, the static background is modeled by the low-rank tucker decomposition, the smooth foreground (moving objects) is modeled by the spatiotemporal continuity, which is enforced by the total variation regularization, and the noise is modeled by the sparsity, which is enforced by the L1 norm. An efficient algorithm based on alternating direction method of multipliers (ADMM) is implemented to solve the proposed model. Extensive experiments on both simulated and real data demonstrate that the proposed method significantly outperforms the state-of-the-art approaches for background/foreground separation in noisy cases.

A document by Caijilahu Bao . Click on the document to view its contents.

Postprint accepted on 30 April 2021 for publication in IEEE International Symposium on Information Theory (ISIT), 2021. (c) 2021 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works. In this paper, we propose a precoded faster-than-Nyquist (FTN) signaling technique for time-domain single-carrier index modulated (IM) symbol transmission. More precisely, eigenvalue decomposition precoding is adopted for the FTN transmission of data bits modulated by single-carrier time-domain IM. While the FTN scheme increases the spectral efficiency and data rate by packing more transmit symbols per block duration than those defined in the Nyquist criterion, time-domain IM works towards the same objective while maintaining symbol sparsity. We analytically derive the constrained capacity of the proposed system. Our simulation results show that the proposed scheme has better bit error ratio (BER) performance over the conventional FTN-IM scheme, particularly for the scenario of a higher packing ratio. In the proposed scheme, $2.5$-dB performance gain is observed at the BER of 10-4, employing the packing ratio of $0.7$ and the roll-off factor of $0.5$ in a channel-uncoded scenario.

Publication reference: https://doi.org/10.1007/978-981-99-9037-5_10 Sudoku puzzles are easily solved using backtracking algorithms. Yet, literature is scattered with various shy and often opaque attempts at using evolutionary algorithms and hybridizing them with known strategies of solving the puzzle. Evolutionary methods are serendipitous in nature, and this paper demonstrates the behaviour of such serendipity under constraints, using visuals that depict the sheer magnitude of the problem space and the nature in which intertwined constraints affect the scope for locating a solution, with the hope that it could inspire a new way of looking at the problem. We propose a method of visualizing the sudoku fitness landscape, the vastness and complexity of even partially brute forcing the puzzle, and a unique method of mutating puzzle states using circular swaps. These insights could potentially serve as a link to comprehend the problem space when designing solutions for vast, multidimensional problems. Additionally, finding the optimal solution for some puzzles was notably harder, compared to puzzles in the same category of given clues. A short investigation was conducted into this phenomenon, which revealed hints that compel us to propose that the direction of research that should be taken, is in discovering more about puzzle states and definitive mathematical properties of the puzzle, rather than merely designing brute-force, stochastic or hybrid approaches of finding solutions.

In this paper, the vacuum reflow soldering technology for semiconductor laser chips in optoelectronic devices was studied and analyzed in a systematic manner. Through the study on the key elements in the reflow soldering process, such as the selection of solders, chamber vacuum, flux, and the pressure applied by the fixture on the chip, this paper focused on exploring the formation mechanism of voids in the solder layer when the device was resoldered. Also, the change in the movement of gas bubbles in the voids with changing reflow oven chamber conditions and its underlying law were analyzed, by preparing 200 C-package semiconductor laser diodes and verifying the reliability and stability of the theoretical analysis through inspection and test aging. which could provide a theoretical basis for the use of the vacuum reflow soldering technology to reduce the void rate in the soldering process of devices.

Deep Learning (DL) has been utilized pervasively in the Internet of Things (IoT). One typical application of DL in IoT is device identification from wireless signals, namely Non-cryptographic Device Identification (NDI). However, learning components in NDI systems have to evolve to adapt to operational variations, such a paradigm is termed as Incremental Learning (IL). Various IL algorithms have been proposed and many of them require dedicated space to store the increasing amount of historical data, and therefore, they are not suitable for IoT or mobile applications. However, conventional IL schemes can not provide satisfying performance when historical data are not available. In this paper, we address the IL problem in NDI from a new perspective, firstly, we provide a new metric to measure the degree of topological maturity of DNN models from the degree of conflict of class-specific fingerprints. We discover that an important cause for performance degradation in IL enabled NDI is owing to the conflict of devices’ fingerprints. Second, we also show that the conventional IL schemes can lead to low topological maturity of DNN models in NDI systems. Thirdly, we propose a new Channel Separation Enabled Incremental Learning (CSIL) scheme without using historical data, in which our strategy can automatically separate devices’ fingerprints in different learning stages and avoid potential conflict. Finally, We evaluated the effectiveness of the proposed framework using real data from ADS-B (Automatic Dependent Surveillance-Broadcast), an application of IoT in aviation. The proposed framework has the potential to be applied to accurate identification of IoT devices in a variety of IoT applications and services. Data and code available at IEEE Dataport (DOI: 10.21227/1bxc-ke87) and \url{https://github.com/pcwhy/CSIL}}.

Automatic Modulation Classification (AMC) receives significant interest in the context of current and future wireless communication systems. Deep learning emerged as a powerful AMC tool, as it allows for the joint learning of discriminative features, and signal classification. However, the optimization of Deep Neural Network (DNN) architectures for AMC is a manual and time-consuming process that requires profound domain knowledge and much effort. Moreover, most proposed solutions focus mainly on classification accuracy, while optimization of network complexity is neglected. In this paper, we propose a novel bi-objective memetic algorithm, BO-NSMA, to search optimal DNN architectures for AMC to maximize classification accuracy and minimize network complexity. We show that BO-NSMA, with a small initial population of six individuals and only ten generations, finds a DNN architecture that outperforms all human-crafted State-of-the-Art (SoA) models. BO-NSMA discovered the first low-complexity Convolutional Neural Network (CNN)-based model, which achieves slightly better performance than costly Recurrent Neural Network (RNN)-based approaches, allowing a 2.8-fold reduction in network complexity with 0.7% performance improvement. Compared to its counterparts from Network Architecture Search (NAS), BO-NSMA finds the best architecture, which achieves up to 18.24% accuracy gain and up to a 78.71-fold reduction in network complexity.

We present analytical physics-based compact models for the Schottky barriers at the interfaces between the organic semiconductor and the source and drain contacts in organic thin-film transistors (TFTs) fabricated in the coplanar and the staggered device architecture, and we illustrate the effect of these Schottky barriers on the current-voltage characteristics of the TFTs. The model for the source barrier explicitly takes into account the field-dependent barrier lowering due to image charges. Potential solutions have been derived by applying the Schwarz-Christoffel transformation, leading to expressions for the electric field at the source barrier and for the contact resistance at the source contact. With regard to the drain barrier, a generic compact-modeling scheme based on the current-voltage characteristics of a barrier-less TFT is introduced that can be applied to any compact dc model. Finally, both models are incorporated into an existing charge-based compact dc model and verified against the results of measurements performed on coplanar and staggered organic TFTs with channel lengths ranging from 0.5 μm to 10.5 μm.

The ubiquity of sensor technology and the Internet of Things prompted us to propose to develop an end-to-end communication architecture for real-time digital dashboards to visualize the anxiety risks of a population during a pandemic, as in the case of COVID-19. Such an architecture can be regarded as the next-generation anxiety risk classification mean for the healthcare industry 4.0 as it will be capable of generating automated and quick actions through the use of analytics on the collected data and predefined thresholds. Based on Internet of Things and wearable healthcare sensors, the proposed end-to-end communication architecture is capable of detecting physiological data related to heart rate, blood pressure, and SPO2, and communicate them to remote cloud servers. Based on this collected data, the centralized dashboard will classify in real time the patients of each geographic region involved according to a specific attribute, namely: normal, mild, moderate, high, severe, or extreme. In addition, we also propose to incorporate the emerging technologies of Space Time Frequency Spreading (STFS) and Space-Time Spreading-Aided Indexed Modulation (STS-IM) for the design of the communication links. It has been found that the integration of STFS and STS-IM promises to reduce the likelihood of data disruption for the proposed architecture.

The reconfigurable intelligent surface (RIS) is an emerging promising candidate technology for the sixth-generation wireless networks, where the element spacing is usually of sub-wavelength. Only limited knowledge, however, has been gained about the spatial-temporal correlation behavior among the elements in an RIS. In this paper, we investigate the spatial-temporal correlation models for an RIS in a wireless communication system. Joint small-scale spatial-temporal correlation functions are provided and analyzed for both ideal isotropic scattering and more practical non-isotropic scattering environments, where the latter is studied via employing an angular distribution derived from real-world millimeter-wave measurements. Furthermore, for the special case of spatial-only correlation under isotropic scattering, an analytical expression is proposed to characterize the spatial degrees of freedom (DoF) for RISs with finite element spacing and aperture sizes in practice. Analytical and numerical results demonstrate that the joint spatial-temporal correlation can be represented by a four-dimensional sinc function under isotropic scattering, while the correlation is generally stronger with more fluctuation and significantly fewer dominant eigenvalues hence smaller DoF for non-isotropic scattering.

Information is mounting exponentially, and the world is moving to hunt knowledge with the help of Big Data. The labelled data is used for automated learning and data analysis which is termed as Machine Learning. Linear Regression is a statistical method for predictive analysis. Gradient Descent is the process which uses cost function on gradients for minimizing the complexity in computing mean square error. This work presents an insight into the different types of Gradient descent algorithms namely, Batch Gradient Descent, Stochastic Gradient Descent and Mini-Batch Gradient Descent, which are implemented on a Linear regression dataset, and hence determine the computational complexity and other factors like learning rate, batch size and number of iterations which affect the efficiency of the algorithm.

This paper presents an extension of the models used to generate fractal traffic flows in high-speed computer networks by means of the formulation of a model that considers the use of one-dimensional chaotic maps. Based on the disaggregation of the temporal series generated, a valid explanation of behavior of the values of Hurst exponent is proposed and the feasibility of their control from the parameters of the proposed model is shown.

Advancements in mobile phones and technologies, have forced the mobile software community to develop smart applications which can help users in general purposes. During the pandemic, people have shifted their focus onto Video chatting and conferencing applications for conducting business, schools, and universities. The current software's available provide several useful functions like live transcripts, allow users to create notes and providing third party application support. But there are certain areas where, video conferencing applications can improve upon primarily in interns of language support. This application will provide a new video conferencing feature which is live translation of speech which could help conduct business and conducting lectures.

Over-sampling is a very effective method to solve the imbalanced problem by generating new synthetic samples for the minority class. But rare over-sampling methods focus on the borderline between classes and only use the linearinterpolation between boundary samples to fill the decision boundary, so not take full use of information in the decision boundary at all. To fill this gap, one novel method named Decision Boundary Computation-based Oversampling is proposed. Firsts, the novel method treats surrounding areas of both boundary majority and minority samples as the decision boundary. Then, compute it’s area belonging to majority class and treat the remained one as the area belonging to the minority class. Thus, the novel method greatly enhances the full use of boundary information and implicitly complements the nature insufficiency of information of minority class at the same time. Finally, new synthetic samples are generated in the partition of decision boundary of minority class. Extensive experiments indicate the good performance of proposed method when compared with other state-of-art methods.

The strict latency constraints of emerging vehicular applications make it unfeasible to forward sensing data from vehicles to the cloud for processing. To shorten network latency, Vehicular fog computing (VFC) moves computation to the edge of the Internet, with the extension to support the mobility of distributed computing entities. In other words, VFC proposes to complement stationary fog nodes co-located with cellular base stations with mobile ones carried by moving vehicles. Previous works of VFC mainly focus on optimizing the assignments of computing tasks among available fog nodes. However, capacity planning, which decides where and how much capacity to deploy, remains an open and challenging issue. The complexity of this problem comes from the mobility of vehicles, the spatio-temporal dynamics of vehicular traffic, and the computing resource demand generated by varying vehicular applications. To solve the above challenges, we propose a data-driven capacity planning framework that optimizes the deployment of stationary and mobile fog nodes to minimize the installation and operational costs under the quality-of-service constraints, taking into account the spatio-temporal variation in computing demand. Through real-world experiments, we analyze the cost efficiency potential of VFC in long term and demonstrate that the performance loss of VFC is below $6\%$ compared to stationary deployment with equal network capacity. We also analyze the impacts of traffic patterns on the potential cost saving. The results show when the traffic density is higher, more operational costs will be saved in the long run due to more dense deployment of mobile fog nodes.

The increased penetration of wind power plants (WPPs) in distribution networks challenges the distribution system operators (DSOs) to improve and optimize networks’ operation. A higher amount of local power production translates to more losses in the network. This paper proposes a deterministic optimization methodology to minimize the losses in distribution networks with WPPs, by exploiting WPPs’ capability to control reactive power in coordination with the on-load tap changers from the MV/HV transformer, avoiding the need for network reinforcements. The principal objective is to optimize the reactive power flow in the network. Measurements from a real distribution network with a large share of controllable WPPs under varying wind and load conditions are used for the study. The benefits and the challenges of the optimization methodology are assessed and discussed with respect to active power losses, voltage profile and reactive power. The results show that with reactive power support from WPPs, network losses are reduced by 4.2 %. Higher loss reductions (up to 19 %) can be achieved through a coordinated action between the WPPs and TSO. Furthermore, it is shown that the distribution network can act as an asset to the transmission network for reactive power support, via actively controlling WPP’s reactive power.