Bandwidth Constraints Research Articles

Event-Triggered Communication Network With Limited-Bandwidth Constraint for Multi-Agent Reinforcement Learning.

Communicating agents with each other in a distributed manner and behaving as a group are essential in multi-agent reinforcement learning. However, real-world multi-agent systems suffer from restrictions on limited bandwidth communication. If the bandwidth is fully occupied, some agents are not able to send messages promptly to others, causing decision delay and impairing cooperative effects. Recent related work has started to address the problem but still fails in maximally reducing the consumption of communication resources. In this article, we propose an event-triggered communication network (ETCNet) to enhance communication efficiency in multi-agent systems by communicating only when necessary. For different task requirements, two paradigms of the ETCNet framework, event-triggered sending network (ETSNet) and event-triggered receiving network (ETRNet), are proposed for learning efficient sending and receiving protocols, respectively. Leveraging the information theory, the limited bandwidth is translated to the penalty threshold of an event-triggered strategy, which determines whether an agent at each step participates in communication or not. Then, the design of the event-triggered strategy is formulated as a constrained Markov decision problem and reinforcement learning finds the feasible and optimal communication protocol that satisfies the limited bandwidth constraint. Experiments on typical multi-agent tasks demonstrate that ETCNet outperforms other methods in reducing bandwidth occupancy and still preserves the cooperative performance of multi-agent systems at the most.

An attitude cooperative control for multi‐spacecraft systems with arbitrary input delay time‐variation rate: A dynamic event‐triggered approach

SummaryA problem of the attitude cooperative control of multi‐spacecraft systems with input time‐varying delay under an undirected graph is addressed in this paper, where the delay does not require any assumption on the continuity of the delay and its time‐variation rate. Considering the communication bandwidth constraint and Zeno behavior, a dynamic event‐triggered scheme is presented, which is more general than some existing event‐triggered schemes with a fixed threshold. By Lyapunov function technology, sufficient conditions of bilinear matrix inequality (BMI) for the existence of controller gain are proposed firstly such that the closed‐loop system is attitude cooperative with a new disturbance attenuation performance. It is also proved that there is a lower bound between two adjacent triggers via the mathematical analytical method, that is, the Zeno phenomenon can be avoided. Finally, the efficacy of the proposed methods is verified via simulations.

Collaborative Semantic Communication for Edge Inference

We study the collaborative image retrieval problem at the wireless edge, where multiple edge devices capture images of the same object from different angles and locations, which are then used jointly to retrieve similar images at the edge server over a shared multiple access channel (MAC). We propose two novel deep learning-based joint source and channel coding (JSCC) schemes for the task over both additive white Gaussian noise (AWGN) and Rayleigh slow fading channels, with the aim of maximizing the retrieval accuracy under a total bandwidth constraint. The proposed schemes are evaluated on a wide range of channel signal-to-noise ratios (SNRs), and shown to outperform the single-device JSCC and the separation-based multiple-access benchmarks. We also propose a channel state information-aware JSCC scheme with attention modules to enable our method to adapt to varying channel conditions.

Characterizing the Performance of Accelerated Jetson Edge Devices for Training Deep Learning Models

Deep Neural Network (DNN) models are becoming ubiquitous in a variety of contemporary domains such as Autonomous Vehicles, Smart cities and Healthcare. They help drones to navigate, identify suspicious activities from safety cameras, and perform diagnostics over medical imaging. Fast DNN inferencing close to the data source is enabled by a growing class of accelerated edge devices such as NVIDIA Jetson and Google Coral which host low-power Graphics Processing Units (GPUs) and Tensor Processing Units (TPUs) along with ARM CPUs in a compact form-factor to offer a superior performance-to-energy ratio. E.g., the NVIDIA Jetson AGX Xavier kit has a 512-core Volta GPU, an 8-core ARM CPU and 32GB LPDDR4x memory, that operates within 65W of power, costs US999 and is smaller than a paperback novel. Recently, there has been a push towards training DNN models on the edge [2]. This is driven by the massive growth in data collected from edge devices in Cyber-Physical Systems (CPS) and Internet of Things (IoT), the need to refresh the models periodically, the bandwidth constraints in moving all this data to Cloud data centers for training, and a heightened emphasis on privacy by retaining data on the edge. This has led to techniques like federated and geo-distributed learning that train DNN models locally on data on an edge device and aggregate them centrally. In this abstract, we summarise and highlight key results from our full paper [5].

Data-Driven Edge Offloading for Wireless Control Systems

As industrial plants embrace modern technologies, such as edge computing and wireless networks, industrial control systems have evolved into multi-tier cyber-physical systems. While traditional local controllers enjoy reliable connectivity to sensors/actuators, they suffer from the limited computation capacity of embedded devices. In contrast, edge servers introduce more computation resources connected to sensors/actuators through wireless networks. Offloading control functions to edge servers presents new opportunities to enhance control performance but also poses critical challenges. As wireless networks have limited bandwidth and varying reliability, it is important to optimize control performance by dynamically offloading a subset of the control functions to edge servers under the bandwidth constraint. Furthermore, the selection of offloaded control functions depends on both the cyber (wireless) and physical states of the wireless control systems. In this paper, we tackle the problem of optimizing the control performance of multiple control loops through dynamic edge offloading. We establish a data-driven model to predict the control performance of each feedback control loop based on its cyber-physical states. We then develop a dynamic edge offloading approach to optimize the overall control performance of a system with multiple feedback control loops while guaranteeing their stability under fluctuating cyber-physical conditions. Finally, we demonstrate the efficacy of the data-driven model and offloading approach in case studies comprising simulations of up to twenty industrial robots.

Two-photon imaging of soliton dynamics

Optical solitary waves (solitons) that interact in a nonlinear system can bind and form a structure similar to a molecule. The rich dynamics of this process have created a demand for rapid spectral characterization to deepen the understanding of soliton physics with many practical implications. Here, we demonstrate stroboscopic, two-photon imaging of soliton molecules (SM) with completely unsynchronized lasers, where the wavelength and bandwidth constraints are considerably eased compared to conventional imaging techniques. Two-photon detection enables the probe and tested oscillator to operate at completely different wavelengths, which permits mature near-infrared laser technology to be leveraged for rapid SM studies of emerging long-wavelength laser sources. As a demonstration, using a 1550 nm probe laser we image the behavior of soliton singlets across the 1800–2100 nm range, and capture the rich dynamics of evolving multiatomic SM. This technique may prove to be an essential, easy-to-implement diagnostic tool for detecting the presence of loosely-bound SM, which often remain unnoticed due to instrumental resolution or bandwidth limitations.

Relay-Assisted Federated Edge Learning: Performance Analysis and System Optimization

In this paper, we study a relay-assisted federated edge learning (FEEL) network under latency and bandwidth constraints. In this network, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">N</i> users collaboratively train a global model assisted by <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">M</i> intermediate relays and one edge server. We firstly propose partial aggregation and spectrum resource multiplexing at the relays in order to improve the communication of the relay-assisted FEEL system. Furthermore, we derive analytical and asymptotic expressions of the system outage probability and convergence rate. For the purpose of improving the system performance, we further optimize the relay-assisted FEEL network by maximizing the number of users who participate in each round of federated learning, through allocation of the wireless bandwidth among users and relays. Specifically, two bandwidth allocation (BA) schemes have been proposed, assuming either instantaneous or statistical channel state information (CSI). Simulations show the advantages of the proposed BA schemes over other benchmarks, regarding the accuracy and convergence rate of the considered relay-assisted FEEL network.

An Efficient On-Chip Data Storage and Exchange Engine for Spaceborne SAR System

Advancements in remote sensing technology and very-large-scale integrated circuit (VLSI) have significantly augmented the real-time processing capabilities of spaceborne synthetic aperture radar (SAR), thereby enhancing terrestrial observational capacities. However, the inefficiency of voluminous data storage and transfer inherent in conventional methods has emerged as a technical hindrance, curtailing real-time processing within SAR imaging systems. To address the constraints of a limited storage bandwidth and inefficient data transfer, this study introduces a three-dimensional cross-mapping approach premised on the equal subdivision of sub-matrices utilizing dual-channel DDR3. This method considerably augments storage access bandwidth and achieves equilibrium in two-dimensional data access. Concurrently, an on-chip data transfer approach predicated on a superscalar pipeline buffer is proposed, mitigating pipeline resource wastage, augmenting spatial parallelism, and enhancing data transfer efficiency. Building upon these concepts, a hardware architecture is designed for the efficient storage and transfer of SAR imaging system data, based on the superscalar pipeline. Ultimately, a data storage and transfer engine featuring register addressing access, configurable granularity, and state monitoring functionalities is realized. A comprehensive imaging processing experiment is conducted via a “CPU + FPGA” heterogeneous SAR imaging system. The empirical results reveal that the storage access bandwidth of the proposed superscalar pipeline-based SAR imaging system’s data efficient storage and transfer engine can attain up to 16.6 GB/s in the range direction and 20.0 GB/s in the azimuth direction. These findings underscore that the storage exchange engine boasts superior storage access bandwidth and heightened data storage transfer efficiency. This considerable enhancement in the processing performance of the entire “CPU + FPGA” heterogeneous SAR imaging system renders it suitable for application within spaceborne SAR real-time processing systems.

Memory Bandwidth Efficient Design for Super-Resolution Accelerators With Structure Adaptive Fusion and Channel-Aware Addressing

State-of-the-art (SOTA) super-resolution (SR) models can generate high-quality images. However, they require a large external memory bandwidth, making it impossible to implement these models on hardware. Although some work has presented different kinds of layer fusion to reduce memory traffic, they can only work on simple model architectures and only consider the feature extraction part. To solve the above issues, this article proposes structure adaptive fusion (SAF) for the feature extraction part to avoid intermediate feature map I/O. This method selects the repetitive structure as the fusion unit and fuses multiple ones to meet buffer size and memory bandwidth constraints, which can deal with different SR models. In addition, we also propose channel-aware addressing for the upscale part to avoid off-chip data transfers. The proposed methods achieve over 90% of memory traffic reduction in all tested SOTA models. Compared to the SOTA fusion method, our approach requires a 52% smaller buffer size and up to 61% lower memory bandwidth for the same number of fusion layers.

Joint optimization of recommendation and caching based on user preference prediction

AbstractThe development of the Internet of things brings exponential growth of wireless traffic, which puts great pressure on the backhaul link. The proactive caching of some contents in the edge device of mobile network can effectively reduce the repeated transmission of the same contents and relieve the burden of the backhaul link. Moreover, the introduction of recommendation mechanisms can reshape user's request and improve cache hit ratio. However, the optimization of recommendation and caching decisions is highly dependent on the users’ preference information for files. Here, a joint optimization algorithm of recommendation and caching based on users’ preference prediction with multiple base stations cooperative caching is proposed. To improve the caching efficiency, the Deep Crossing model is adopted to predict users’ preferences. Under the constraints of cache capacity, recommendation quantity and bandwidth, an optimization problem to minimize the total transmission delay of the system is formulated. Then, the NP‐hardness of the proposed optimization problem is proved and it is decoupled it into three sub‐problems, namely recommendation, user access and caching optimization sub‐problems. Simulation results show that the proposed algorithm can effectively reduce the total transmission delay of the system.

Distributed discrete-time optimization over directed networks: A dynamic event-triggered algorithm

Distributed optimization has a wide variety of applications, such as economic dispatch in power systems, social welfare maximization, optimal power flow, where distributed nodes work cooperatively to achieve a minimized global objective function, where the information exchanges among distributed and networked nodes play an important role. However, such information exchanges often suffer from inevitable energy and bandwidth constraints due mainly to their distributed and networked nature. Event-triggered distributed optimization provides an effective means to address this issue. However, designing an event-triggered distributed optimization algorithm with both fast convergence and resource efficiency is challenging. In this paper, a distributed zero-gradient-sum algorithm with a dynamic event-triggered scheme is presented for distributed discrete-time optimization. The triggering threshold is composed of an exponential decaying term and a positive dynamic auxiliary variable. The linear convergence is established theoretically. Finally, this algorithm is applied to an economic dispatch problem to verify the theoretical findings and to demonstrate the effectiveness and superiority of the proposed algorithm.

IFogRep: An intelligent consistent approach for replication and placement of IoT based on fog computing

Recently, the Internet of Things (IoT) has played key roles in every aspect of daily routine tasks. IoT applications suffer from higher latency and limited network bandwidth in the internet core. However, due to bandwidth and resource constraints, the resultant massive data traffic triggers storage overhead, high latency, and network bottleneck problems. Therefore, cloud computing and fog computing are introduced as popular IoT assistance. In fact, using a fog layer with such a highly dynamic environment raises a placement problem that needs to identify the optimal fog node that can be assigned to place the sensed data there. However, using a single copy of the datum in a single node can increase the concurrent access and maximize the network latency. Therefore, replication is considered a known technique to improve availability by publishing many replicas to many sites. However, using replication causes a degree of temporal data inconsistency and requires the application to be tolerant. This work proposes a novel intelligent solution for dynamic consistent replication and optimal placement of IoT using fog computing. The results show the ability of the current work to reduce the system latency and cost of replica access and increase the data availability.

Learning-driven lossy image compression: A comprehensive survey

In the field of image processing and computer vision (CV), machine learning (ML) architectures are widely used. Image compression problems can be solved using convolutional neural networks (CNNs). As a result of bandwidth and memory constraints, compression of images is a necessity. There are three types of information found in images: useful, redundant, and irrelevant. In this survey, we will discuss how ML is used to compress lossy images. Firstly, we describe the background of lossy image compression. Next, we classify ML-based image compression frameworks into subgroups based on their architectures. Auto-encoders (AEs), variational auto-encoders (VAEs), CNNs, recurrent neural networks (RNNs), long short-term memories (LSTMs), gated recurrent units (GRUs), generative adversarial networks (GANs), transformers, principal component analysis (PCA) and fuzzy means clustering are among these subgroups. By analyzing learning-driven image compression frameworks, we present pros and cons of each subgroup. Lastly, we outline several research gaps and future research directions in the field of ML-based image compression.

A smart energy IoT model based on the Itsuku PoW technology

Based on the blockchain + software defined network(SDN) technology, we explore and research a new smart energy networks IoT architecture model of distributed trust to solve the “stuck neck” problems such as bandwidth constraints and trust obstruction of the current centralized network architecture of smart energy networks. In order to achieve efficient data flow and sharing and co-governance. Based on the MTP-Argon2 hash function algorithm in the Itsuku PoW technical solution, we add computing power optimization and correction multi-objective optimization operations, and use the network control right separation technology of the FS-Open Security SDN model to establish local database access policies and storage credentials. The traditional hardware devices are separated from the centralized network architecture of the smart energy network into a hybrid network architecture model of distributed trust. The results show that the distributed trust hybrid network architecture model achieves low latency and circumvents network bandwidth limitations, reflecting the energy efficiency gain and security stability of the model, as well as network load robustness.

Improving the ranging performance of chaos LiDAR.

Chaos lidar has gained significant attention due to its high spatial resolution, natural anti-interference capability, and confidentiality. However, constrained by the power of the chaos laser, the sensitivity of the linear detector, and the hardware bandwidth, chaos lidar is greatly restricted in the application of long-distance target detection and imaging. To overcome these constraints, we propose a novel, to the best of our knowledge, chaos lidar based on Geiger mode avalanched photodetectors (GM-APDs) in a previous study called chaos single-photon (CSP) lidar. In this paper, we compare the CSP lidar with the linear mode chaos lidars by combining with lidar equation. Regarding the ranging principle, the CSP lidar is fully digital and breaks through the constraints of a detector's bandwidth and ADC's sampling rate. The simulation results indicate that the detection range of the CSP lidar is approximately 35 times and 8 times greater than that of a continuous-wave chaos lidar and pulsed chaos lidar, respectively. Although the detection accuracy of the CSP lidar is only at the centimeter level and is lower than the linear mode chaos lidars, its consumption of storage resources and power is greatly reduced due to 1-bit quantization in the GM-APD. Additionally, we investigate the impact of GM-APD parameters on the signal-to-noise ratio (SNR) of the CSP lidar system and demonstrate that the dead time difference between GM-APDs has a negligible effect. In conclusion, we present and demonstrate a new chaos lidar system with a large detection range, high SNR, low storage resources and power consumption, and on-chip capability.

A Novel Memory-Based Scheduling Protocol for Networked Control Systems Under Stochastic Attacks and Bandwidth Constraint

A Novel Memory-Based Scheduling Protocol for Networked Control Systems Under Stochastic Attacks and Bandwidth Constraint

Security-Enhanced Serial Communications

Industrial Control Systems (ICS) are widely used by critical infrastructure and are ubiquitous in numerous industries including telecommunications, petrochemical, and manufacturing. ICS are at a high risk of cyber attack given their internet accessibility, inherent lack of security, deployment timelines, and criticality. A unique challenge in ICS security is the prevalence of serial communication buses and other non-TCP/IP communications protocols. The communication protocols used within serial buses often lack authentication and integrity protections, leaving them vulnerable to spoofing and replay attacks. The bandwidth constraints and prevalence of legacy hardware in these systems prevent the use of modern message authentication and integrity techniques, such as those provided by Transport Layer Security (TLS). Our approach seeks to address the challenges of providing authentication to serial communications while keeping the existing serial communications bus in place and adding minimal hardware. Here we demonstrate the integration of field-programmable gate arrays (FPGAs) on the bus to inject message authentication codes into an out-of-band (OOB) communications channel, leveraging Power Line Communication (PLC) techniques. Our results show that simple solutions can be integrated into existing serial communications to provide necessary security features in critically important systems. The presented solution is inexpensive, scales well, is modular and extensible, and has low temporal overhead. Providing authenticated serial communications for critical infrastructure systems will improve their ability to resist cyber attacks.

Designing Globally Time-Optimal Entangling Gates Using Geometric Space Curves

High-fidelity entangling gates are essential for quantum computation. Currently, most approaches to designing such gates are based either on simple, analytical pulse waveforms or on ones obtained from numerical optimization techniques. In both cases, it is typically not possible to obtain a global understanding of the space of waveforms that generate a target gate operation, which can make it challenging to identify globally time-optimal waveforms that respect amplitude and bandwidth constraints. Here, we show that in the case of weakly coupled qubits, it is possible to find all pulses that implement a target entangling gate in near-minimal time. We do this by mapping quantum evolution onto geometric space curves. We derive the minimal conditions these curves must satisfy in order to guarantee a gate with a desired entangling power is implemented. Pulse waveforms are extracted from the curvatures of these curves. We illustrate our method by designing fast, cnot-equivalent entangling gates for silicon quantum dot spin qubits with fidelities exceeding 99 %. We show that fidelities can be further improved while maintaining low bandwidth requirements by using geometrically derived pulses as initial guesses in numerical optimization routines.

Double-event-triggered cooperative maximum correntropy estimation over wireless sensor networks

This paper is concerned with the distributed state estimation problem for a class of multi-sensor network systems with bandwidth constraints, measurement loss, and shot noises. To tackle these issues, the double-event-triggered cooperative maximum correntropy estimation algorithm is proposed following the proposed double-event-triggered communication mechanism and maximum correntropy criterion. With such an algorithm, measurement errors between adjacent events in the first-event trigger are applied to determine whether the sensor node needs to transmit data, and then current innovation samples in the second-event trigger are to detect measurement anomalies. Compared to existing approaches, the proposed algorithm can significantly reduce communication costs and improve sensors’ service life, accuracy and robustness. A simulation case-study concerning moving target tracking with multiple unmanned aerial vehicles (multi-UAVs) is worked out to demonstrate the effectiveness of the considered algorithm.

Future of Drug Discovery: The Synergy of Edge Computing, Internet of Medical Things, and Deep Learning

The global spread of COVID-19 highlights the urgency of quickly finding drugs and vaccines and suggests that similar challenges will arise in the future. This underscores the need for ongoing efforts to overcome the obstacles involved in the development of potential treatments. Although some progress has been made in the use of Artificial Intelligence (AI) in drug discovery, virologists, pharmaceutical companies, and investors seek more long-term solutions and greater investment in emerging technologies. One potential solution to aid in the drug-development process is to combine the capabilities of the Internet of Medical Things (IoMT), edge computing (EC), and deep learning (DL). Some practical frameworks and techniques utilizing EC, IoMT, and DL have been proposed for the monitoring and tracking of infected individuals or high-risk areas. However, these technologies have not been widely utilized in drug clinical trials. Given the time-consuming nature of traditional drug- and vaccine-development methods, there is a need for a new AI-based platform that can revolutionize the industry. One approach involves utilizing smartphones equipped with medical sensors to collect and transmit real-time physiological and healthcare information on clinical-trial participants to the nearest edge nodes (EN). This allows the verification of a vast amount of medical data for a large number of individuals in a short time frame, without the restrictions of latency, bandwidth, or security constraints. The collected information can be monitored by physicians and researchers to assess a vaccine’s performance.