Massive Interconnection Research Articles

Design of Network-on-Chip-Based Restricted Coulomb Energy Neural Network Accelerator on FPGA Device.

Sensor applications in internet of things (IoT) systems, coupled with artificial intelligence (AI) technology, are becoming an increasingly significant part of modern life. For low-latency AI computation in IoT systems, there is a growing preference for edge-based computing over cloud-based alternatives. The restricted coulomb energy neural network (RCE-NN) is a machine learning algorithm well-suited for implementation on edge devices due to its simple learning and recognition scheme. In addition, because the RCE-NN generates neurons as needed, it is easy to adjust the network structure and learn additional data. Therefore, the RCE-NN can provide edge-based real-time processing for various sensor applications. However, previous RCE-NN accelerators have limited scalability when the number of neurons increases. In this paper, we propose a network-on-chip (NoC)-based RCE-NN accelerator and present the results of implementation on a field-programmable gate array (FPGA). NoC is an effective solution for managing massive interconnections. The proposed RCE-NN accelerator utilizes a hierarchical-star (H-star) topology, which efficiently handles a large number of neurons, along with routers specifically designed for the RCE-NN. These approaches result in only a slight decrease in the maximum operating frequency as the number of neurons increases. Consequently, the maximum operating frequency of the proposed RCE-NN accelerator with 512 neurons increased by 126.1% compared to a previous RCE-NN accelerator. This enhancement was verified with two datasets for gas and sign language recognition, achieving accelerations of up to 54.8% in learning time and up to 45.7% in recognition time. The NoC scheme of the proposed RCE-NN accelerator is an appropriate solution to ensure the scalability of the neural network while providing high-performance on-chip learning and recognition.

Direct visualization of domain wall pinning in sub-100 nm 3D magnetic nanowires with cross-sectional curvature.

The study of 3D magnetic nanostructures has uncovered rich phenomena including the stabilization of topological spin textures using nanoscale curvature, controlled spin-wave emission, and novel ground states enabled by collective frustrated interactions. From a technological perspective, 3D nanostructures offer routes to ultrahigh density data storage, massive interconnectivity within neuromorphic devices, as well as mechanical induction of stem cell differentiation. However, the fabrication of 3D nanomagnetic systems with feature sizes down to 10 nm poses a significant challenge. Here we present a means of fabricating sub-100 nm 3D ferromagnetic nanowires, with both cross-sectional and longitudinal curvature, using two-photon lithography at a wavelength of 405 nm, combined with conventional deposition. Nanostructures with lateral features as low as 70 nm can be rapidly and reproducibly fabricated. A range of novel domain walls, with anti-vortex textures and hybrid vortex/anti-vortex textures are enabled by the cross-sectional curvature of the system, as demonstrated by micromagnetic simulations. Magnetic force microscopy experiments in an externally applied magnetic field are used to image the injection and pinning of domain walls in the 3D magnetic nanowire. At specific field values, domain walls are observed to hop from trap to trap, providing a direct means to probe the local energy landscape.

Design Scheme of Distributed Energy Resources Interconnection Oriented the VPP System

At present, the virtual power plant is still in the stage of theoretical research and early pilot development. It has not yet formed a systematic and comprehensive Distributed Energy Resource (DER) modeling, market participation mechanism, operational risk management, operational optimization decision-making method, and systematic operating platform. Given the massive DERs interconnection requirements of the virtual power plant system, this paper proposes a design scheme for the DER interconnection protocol, which solves the problems of the existing application layer protocol, such as high overhead, poor security, and low efficiency of massive interconnection, and supports the safe and stable operation of power grid enterprises and virtual power plant systems.

Point-to-Multipoint Coherent Architecture with Joint Resource Allocation for B5G/6G Fronthaul

With the increasing wireless data and massive interconnections in B5G/6G, demands for an intelligent, flexible, and cost-effective fronthaul network have emerged to provide services for a ubiquitous mobile society. This calls for disruptive innovations in both the architecture and transmission technology. In this article, we introduce a point-to-multipoint coherent architecture as a promising solution, and analyze its advantages and challenges compared to other fronthaul solutions. Based on this architecture, we propose a novel joint resource allocation framework that enables the fronthaul to intelligently self-manage and self-optimize in real time. In addition, we propose a low-complexity joint resource allocation algorithm for different scenarios in this framework. Simulations show that the proposed framework and algorithm can effectively utilize the limited resources and improve the performance of the fronthaul.

Comb-based photonic neural population for parallel and nonlinear processing

It is believed that neural information representation and processing relies on the neural population instead of a single neuron. In neuromorphic photonics, photonic neurons in the form of nonlinear responses have been extensively studied in single devices and temporal nodes. However, to construct a photonic neural population (PNP), the process of scaling up and massive interconnections remain challenging considering the physical complexity and response latency. Here, we propose a comb-based PNP interconnected by carrier coupling with superior scalability. Two unique properties of neural population are theoretically and experimentally demonstrated in the comb-based PNP, including nonlinear response curves and population activities coding. A classification task of three input patterns with dual radio-frequency (RF) tones is successfully implemented in a time-efficient manner, which allows the comb-based PNP to make effective use of the ultra-broad bandwidth of photonics for parallel and nonlinear processing.

Reconfigurable optoelectronic memristor for in-sensor computing applications

Inspired by human brain and visual system, optoelectronic memristors-based neuromorphic computing has attracted the interests of researchers to overcome the limitation of traditional von Neumann architecture. With advantages of highly parallel computing and massive interconnection, the optical memristors could construct light-inspired artificial neural network for neuromorphic computing tasks. Besides, nonvolatile optoelectronic memristors provide a promising path for reconfigurable logic operations, greatly promoting the development of novel in-memory computing technology. In this work, the photoelectric perception, storage and in situ computing functions were integrated in optoelectronic memristors array, which could greatly decrease the footprint of multifunctional device and improve the work efficiency of chip. The neuromorphic computing capability of the photonic memristors was verified using face images of different people with accuracy of 86.7%. Moreover, with the advantages of photoelectric cooperative modulation, the reconfigurable logic functions including “AND” and “OR” were achieved by optoelectronic memristors. The present results demonstrate the attractive bio-inspired in-sensor computing behaviors of the optoelectronic memristors, opening up potential applications of optoelectronic memristors in next-generation reconfigurable sensing-memory-computing integrated paradigms.

A Review of Low-Voltage Renewable Microgrids: Generation Forecasting and Demand-Side Management Strategies

It is expected that distribution power systems will soon be able to connect a variety of microgrids from residential, commercial, and industrial users, and thus integrate a variety of distributed generation technologies, mainly renewable energy sources to supply their demands. Indeed, some authors affirm that distribution networks will propose significant changes as a consequence of this massive integration of microgrids at the distribution level. Under this scenario, the control of distributed generation inverters, demand management systems, renewable resource forecasting, and demand predictions will allow better integration of such microgrid clusters to decongest power systems. This paper presents a review of microgrids connected at distribution networks and the solutions that facilitate their integration into such distribution network level, such as demand management systems, renewable resource forecasting, and demand predictions. Recent contributions focused on the application of microgrids in Low-Voltage distribution networks are also analyzed and reviewed in detail. In addition, this paper provides a critical review of the most relevant challenges currently facing electrical distribution networks, with an explicit focus on the massive interconnection of electrical microgrids and the future with relevant renewable energy source integration.

Joint Optimization of Cooperative Edge Caching and Radio Resource Allocation in 5G-Enabled Massive IoT Networks

The fifth-generation of wireless communication (5G) is a promising paradigm toward massive interconnectivity within Internet-of-Things (IoT) networks. However, because the data traffic throughput sharply increases with the number of IoT devices, a tremendous burden on the backhaul links and core networks results. With this in mind, mobile edge caching is an effective method that can relieve stress of the backhaul links, while decreasing the service latency. The purpose of this study is to analyze the problem of jointly optimizing cooperative edge caching and radio resource allocation in 5G-enabled massive IoT networks. For that, a joint optimization long-term nonlinear integer programming problem is posed. This class of problems is known to be NP-hard; thus, to reduce the problem complexity, a divide and conquer scheme will be applied—the task at hand will be divided into two subproblems: 1) cooperative edge caching and 2) radio resource allocation. The cooperative edge caching subproblem is formulated as a constrained Markov decision process. Herein, a deep reinforcement learning method to optimize the caching decisions for all the edge nodes. Then, based on the resulting optimal caching decisions, the radio resource allocation subproblem for each edge node is posed as an NLIP problem, and an improved branch-and-bound method is proposed to yield the optimal radio resource allocation decisions for each edge node. Extensive simulations were performed to confirm that the proposed methods have the capability of enhancing the content caching hit ratio, while lessening the content retrieving delays for 5G-enabled massive IoT networks—improving over various baseline algorithms.

On deep reinforcement learning security for Industrial Internet of Things

The Industrial Internet of Things (IIoT), also known as Industry 4.0, empowers manufacturing and production processes by leveraging automation and Internet of Things (IoT) technologies. In IIoT, the information communication technologies enabled by IoT could greatly improve the efficiency and timeliness of information exchanges between both vertical and horizontal system integrations. Likewise, machine learning algorithms, particularly Deep Reinforcement Learning (DRL), are viable for assisting in automated control of complex IIoT systems, with the support of distributed edge computing infrastructure. Despite noticeable performance improvements, the security threats brought by massive interconnections in IoT and the vulnerabilities of deep neural networks used in DRL must be thoroughly investigated and mitigated before widespread deployment. Thus, in this paper we first design a DRL-based controller that could be deployed at edge computing server to enable automated control in an IIoT context. We then investigate malicious behaviors of adversaries with two attacks: (i) function-based attacks that can be launched during training phase and (ii) performance-based attacks that can be launched after training phase, to study the security impacts of vulnerable DRL-based controllers. From the adversary’s perspective, maximum entropy Inverse Reinforcement Learning (IRL) is used to approximate a reward function through observation of system trajectories under the control of trained DRL-based controllers. The approximated reward function is then used to launch attacks by the adversary against the Deep Q Network (DQN)-based controller. Via simulation, we evaluate the impacts of our two investigated attacks, finding that attacks are increasingly successful with increasing accuracy of the control model. Furthermore, we discuss some tradeoffs between control performance and security performance of DRL-based IIoT controllers, and outline several future research directions to secure machine learning use in IIoT systems.

Analiza vrednosti kao mehanizam za redukovanje kompleksnosti izbora sistema resursa za agilna/virtuelna preduzeća u kontekstu iIndustrije 4.0

The selection of the resources system (SRS) is an important element in the integration/project of Agile/Virtual Enterprises (A/V E) because its performance is dependent of this selection, and even of its creation. However, it remains a difficult matter to solve because is still a very complex and uncertain problem. We propose that using Value Analysis (VA) in the pre-selection of resources phase represents a significant improvement of the SRS process. The current literature fails to formally address the pre-selection phase and none of the resource selection models incorporate the resources value and its consequence in the complexity of the selection process. Whereby, ours developed model with VA constitutes an innovative approach towards greater sustainability in the configuration of A/V E in the context of Industry 4.0, where a massive interconnection among enterprises is expected and consequently the increase of the selection process complexity. After the construction of a demonstrator tool for a set of the problem formulations, this paper verifies by computational results the thesis regarding the benefits of applying VA to the SRS process: VA reduces the complexity of the SRS process, even ensuring that the final system of resources achieve higher quality/value grade.

6G Massive Radio Access Networks: Key Applications, Requirements and Challenges

Driven by the emerging use cases in massive access future networks, technological advancements and evolutions are needed for wireless communications beyond the fifth-generation (5G) networks. In particular, we envisage that the upcoming sixth-generation (6G) networks with numerous devices will demand extremely high-performance interconnections, even under strenuous scenarios such as diverse mobility, extreme density, and dynamic environments. To cater to such a demand, an investigation of flexible and sustainable radio access network (RAN) techniques to support highly diverse requirements and massive connectivity is of utmost importance. To this end, this paper first presents the key driving applications for 6G, including smart cities and factories, unmanned aerial vehicles, and multi-dimensional sensing services, which necessitate the transformation of existing RAN techniques to achieve the key performance indicators required for 6G networks. In particular, we investigate the flexibility, massive interconnectivity, and energy efficiency aspects of 6G RANs and provide an in-depth discussion on the potential key enabling technologies. Last but not least, we present an artificial-intelligence-assisted network architecture for flexible 6G massive RANs with energy harvesting.

BTMPP: Balancing Trust Management and Privacy Preservation for Emergency Message Dissemination in Vehicular Networks

As a potential application field of the sixth-generation (6G) communication technology and a promising part of massive Internet of Things (IoT), vehicular networks have attracted considerable attention from both academia and industry in recent years, where the cooperative safety applications are a significant branch. It is widely acknowledged that 6G is able to provide high-throughput and low-latency wireless communication capability for vehicular networks, support massive interconnectivity in vehicular networks with diverse service requirements, and significantly improve the performance of vehicular networks. Both trust management and privacy preservation play significant roles in vehicular networks, and there exists a tradeoff between them. For providing a satisfactory solution to balance the trust management and privacy preservation in vehicular networks, we put forward a novel scheme named as BTMPP (which is able to provide both the precise trust management and strong conditional privacy preservation simultaneously) in this article by leveraging the famous bloom filter (BF)-based private set intersection (PSI) technology. Furthermore, the theoretical analysis for the correctness, strong conditional privacy preservation capability, strong robustness, precise trust management, and the other features is detailed, and a series of simulations are conducted. The results reveal that the proposed scheme significantly outperforms the existing schemes in several aspects.

Percolation with plasticity for neuromorphic systems

We develop a theory of percolation with plasticity media (PWPs) rendering properties of interest for neuromorphic computing. Unlike the standard percolation, they have multiple (N ≫ 1) interfaces and exponentially large number (N!) of conductive pathways between them. These pathways consist of non-ohmic random resistors that can undergo bias induced nonvolatile modifications (plasticity). The neuromorphic properties of PWPs include: multi-valued memory, high dimensionality and nonlinearity capable of transforming input data into spatiotemporal patterns, tunably fading memory ensuring outputs that depend more on recent inputs, and no need for massive interconnects. A few conceptual examples of functionality here are random number generation, matrix-vector multiplication, and associative memory. Understanding PWP topology, statistics, and operations opens a field of its own calling upon further theoretical and experimental insights.

Complexity of Deep Convolutional Neural Networks in Mobile Computing

Neural networks employ massive interconnection of simple computing units called neurons to compute the problems that are highly nonlinear and could not be hard coded into a program. These neural networks are computation-intensive, and training them requires a lot of training data. Each training example requires heavy computations. We look at different ways in which we can reduce the heavy computation requirement and possibly make them work on mobile devices. In this paper, we survey various techniques that can be matched and combined in order to improve the training time of neural networks. Additionally, we also review some extra recommendations to make the process work for mobile devices as well. We finally survey deep compression technique that tries to solve the problem by network pruning, quantization, and encoding the network weights. Deep compression reduces the time required for training the network by first pruning the irrelevant connections, i.e., the pruning stage, which is then followed by quantizing the network weights via choosing centroids for each layer. Finally, at the third stage, it employs Huffman encoding algorithm to deal with the storage issue of the remaining weights.

A Trustworthy Privacy Preserving Framework for Machine Learning in Industrial IoT Systems

Industrial Internet of Things (IIoT) is revolutionizing many leading industries such as energy, agriculture, mining, transportation, and healthcare. IIoT is a major driving force for Industry 4.0, which heavily utilizes machine learning (ML) to capitalize on the massive interconnection and large volumes of IIoT data. However, ML models that are trained on sensitive data tend to leak privacy to adversarial attacks, limiting its full potential in Industry 4.0. This article introduces a framework named PriModChain that enforces privacy and trustworthiness on IIoT data by amalgamating differential privacy, federated ML, Ethereum blockchain, and smart contracts. The feasibility of PriModChain in terms of privacy, security, reliability, safety, and resilience is evaluated using simulations developed in Python with socket programming on a general-purpose computer. We used Ganache_v2.0.1 local test network for the local experiments and Kovan test network for the public blockchain testing. We verify the proposed security protocol using Scyther_v1.1.3 protocol verifier.

Power flow model with flat frequency control and load characteristic for voltage stability indices

The growing of power systems with massive interconnection becomes more complicated for reliable and economical operation under dynamic as well as steady-state operating conditions. Therefore, it is desirable to develop techniques for evaluating the voltage stability condition with the implementation of the voltage stability index and the minimum singular value of the system. This paper proposes an approach a model of power flow incorporating the generator controls with considering load characteristics of flat frequency control (FFC) scheme in large interconnected power systems. Results on a 24-Bus EHV practical system is presented for illustration of power flow solution.

A context-aware encryption protocol suite for edge computing-based IoT devices

Heterogeneous devices are connected with each other through wireless links within a cyber physical system. These devices undergo resource constraints such as battery, bandwidth, memory and computing power. Moreover, the massive interconnections of these devices result in network latency and reduced speed. Edge computing offers a solution to this problem in which devices transmit the preprocessed actionable data in a formal way, resulting in reduced data traffic and improved speed. However, to provide the same level of security to each piece of information is not feasible due to limited resources. In addition, not all the data generated by Internet of things devices require a high level of security. Context-awareness principles can be employed to select an optimal algorithm based on device specifications and required information confidentiality level. For context-awareness, it is essential to consider the dynamic requirements of data confidentiality as well as device available resources. This paper presents a context-aware encryption protocol suite that selects optimal encryption algorithm according to device specifications and the level of data confidentiality. The results presented herein clearly exhibit that the devices were able to save 79% memory consumption, 56% battery consumption and 68% execution time by employing the proposed context-aware encryption protocol suite.

CONTEXT-AWARE ONTOLOGY AND WEB SERVICES DISCOVERY FOR DISTRIBUTED EMBEDDED REAL-TIME SYSTEMS

Distributed Embedded Real-Time Systems (DERTS) consists of hundreds of interconnected devices, typically small and wirelessly-connected, which are designed to work for a long period of time. The massive interconnection of devices and the usage of heterogeneous languages, operating platforms and data standards make DERTS a competitive and complex system. In addition, the DERTS program is setting a trend of moving away from centralized, high-cost products towards lower cost and high volume products. In this regard, there is nothing more natural than considering the use of Service Oriented Architecture (SOA) to assist in the development of DERTS. This is because SOA enables different devices to exchange data regardless of issues of complexity. Moreover, context-awareness, which is widely studied for DERTS, also plays an important role for effective communication among devices. Thus, to build service-based DERTS while managing the complexity, context-aware ontologies are the best solution. In this paper, we developed a context-aware ontology for DERTS which is known as ConOntDERTS. To evaluate ConOntDERTS, we used two methods. In the first method, a criteria-based ontology evaluation was used; while in the second method, a survey was conducted to show that the results produced by ConOntDERTS were almost the same as human perception. Results of the evaluation show the consistency and feasibility of our ontology and the statistical test results show that ConOntDERTS ontology can produce consistent results with human perception.

5G Message Service System Based on Artificial Immune Dynamic Adaptive Mechanism

When facing massive interconnection scenarios of terminals, 5G message service system needs to support large-scale information access and management and provides the efficient and reliable signal control to scheduling a variety of wireless access networks. In this paper, artificial immune theory is applied to the 5G message services. The simulating the immune response mechanism, a semi-distributed dynamic adaptive immune network architecture was proposed. The dynamic learning mechanism and immune memory mechanism of the detector were established, which improved the 5G edge computing capability and reduced the load of the 5G core network. The concept of the immune message distribution system (IMDS) is proposed, which uses clonal selection algorithm to classify and clone the message header, and combines with the affirmative selection algorithm to carry out high-frequency variation on the message body. On the premise of ensuring the diversity of antibodies, the space consumption problem of the hash-map algorithm is effectively solved, and the efficiency of message distribution is improved. The comparison of simulation results shows that the IMDS improves the message recognition ability and adaptive ability of the 5G system and proves the feasibility of semi-distributed immune dynamic learning mechanism.

Superconducting Optoelectronic Circuits for Neuromorphic Computing

Neural networks have proven effective for solving many difficult computational problems. Implementing complex neural networks in software is very computationally expensive. To explore the limits of information processing, it will be necessary to implement new hardware platforms with large numbers of neurons, each with a large number of connections to other neurons. Here we propose a hybrid semiconductor-superconductor hardware platform for the implementation of neural networks and large-scale neuromorphic computing. The platform combines semiconducting few-photon light-emitting diodes with superconducting-nanowire single-photon detectors to behave as spiking neurons. These processing units are connected via a network of optical waveguides, and variable weights of connection can be implemented using several approaches. The use of light as a signaling mechanism overcomes fanout and parasitic constraints on electrical signals while simultaneously introducing physical degrees of freedom which can be employed for computation. The use of supercurrents achieves the low power density necessary to scale to systems with enormous entropy. The proposed processing units can operate at speeds of at least $20$ MHz with fully asynchronous activity, light-speed-limited latency, and power densities on the order of 1 mW/cm$^2$ for neurons with 700 connections operating at full speed at 2 K. The processing units achieve an energy efficiency of $\approx 20$ aJ per synapse event. By leveraging multilayer photonics with deposited waveguides and superconductors with feature sizes $>$ 100 nm, this approach could scale to systems with massive interconnectivity and complexity for advanced computing as well as explorations of information processing capacity in systems with an enormous number of information-bearing microstates.