Embedding Machine Learning for Resource-Constrained Computing Platforms

A special issue of Electronics (ISSN 2079-9292). This special issue belongs to the section "Computer Science & Engineering".

Deadline for manuscript submissions: closed (31 January 2024) | Viewed by 4536

Special Issue Editors


E-Mail Website
Guest Editor
Institute for Advanced Simulation (IAS), Jülich Supercomputing Centre (JSC), Wilhelm-Johnen-Straße, 52425 Jülich, Germany
Interests: large-scale hardware architectures for the exascale and exaflop operations; HW/SW codesign of high-performance systems, focusing on chip design; heterogeneous architectures (i.e., FPGA) for AI and modular design
Special Issues, Collections and Topics in MDPI journals

E-Mail Website
Guest Editor
Advanced Computing, Photonics and Electromagnetics (CPE) group at Fondazione LINKS, 10138 Turin, Italy
Interests: high-performance computing; cloud computing; quantum computing and applications; hardware accelerator design over FPGAs; evolutionary algorithms and their applications
Special Issues, Collections and Topics in MDPI journals

E-Mail Website
Guest Editor
Department of Digitalization, Copenhagen Business School, Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Interests: data security and privacy; blockchain; machine learning; cloud/fog computing; big data and high-performance/throughput computing
Special Issues, Collections and Topics in MDPI journals

E-Mail Website
Guest Editor
Graduate School of Engineering, the University of Tokyo, Tokyo 113-8656, Japan
Interests: embedded systems; VLSI design methodology

E-Mail Website
Guest Editor
Faculty of Electrical Engineering and Information Technology, Ruhr-Universität Bochum, Bochum, Germany
Interests: embedded systems; low-power; reliability; design automation; FPGA; approximate computing; emerging technologies; logic synthesis

Special Issue Information

Dear Colleagues,

With the growth of computational capability and data, machine learning (ML) applications have become ubiquitous in our personal lives (smart home, smart devices). They are also widely popular in the scientific (medical) and industrial domains. Broadly, ML models are employed either to predict behaviour or extract some meaningful insights from data. The autonomous-X* domain has been broadly employed for object detection/classification, pattern recognition, and natural language processing.

IoT devices also play an essential part in collecting context-specific data, which ML models can further process to achieve higher performance (throughput, lower latency, and scalability). At lower energy cost, application-specific accelerators and complex programming frameworks (with compiler support such as MLIR with TensorFlow) are also being researched and developed. Overall, the scientific community and industry have an increasing interest in making hardware intelligent by porting ML and DL (deep learning being a subset of ML domain) models. Unlike legacy computing equipment (e.g., servers, desktop computers), embedded systems are more severely constrained by additional factors. Such factors include limited power consumption, time-bound processing, and limited scalability. In such scenarios, ML/DL applications can be manifold: from designing embedded accelerators for ML/DL algorithms to ML/DL applications optimising the operations of a given embedded system to ML algorithms for specific scientific/industrial applications leveraging embedded systems.

This Special Issue will cover innovative, holistic approaches to the design of ML/DL algorithms focusing on the convergence between embedded systems and the ML domain. These include (but are not limited to):

  • Custom accelerators for low-powered embedded systems.
  • ML/DL algorithms supporting an embedded system’s (complex) design phases.
  • ML/DL applications leveraging embedded (low-power/distributed) systems.
  • Apply ML/DL models to test the data path.
  • Embedded ML/DL models to add security modules to embedded systems.
  • Methods and tools for parallel large-scale ML/DL applications.
  • Adding ML/DL application support to edge/fog devices.
  • Porting distributed deep learning into embedded systems.
  • Building embedded systems to support hyperparameter optimisation.
  • Incorporation of predictive models to improve the performance of scientific/industrial applications.
  • Building tools and infrastructure to improve the usability of ML/DL models in scientific/industrial applications.
  • Frameworks for optimising HPC ecosystems for embedding ML/DL models.
  • Porting ML compilers into embedded platforms.
  • Architectures for approximate computing.

Dr. Antoni Portero
Dr. Alberto Scionti
Dr. Somnath Mazumdar
Prof. Dr. Fujita Masahiro
Prof. Dr. Akash Kumar
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Electronics is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2400 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.

Keywords

  • low power
  • high performance
  • machine learning
  • deep learning
  • embedded systems
  • hardware accelerators

Benefits of Publishing in a Special Issue

  • Ease of navigation: Grouping papers by topic helps scholars navigate broad scope journals more efficiently.
  • Greater discoverability: Special Issues support the reach and impact of scientific research. Articles in Special Issues are more discoverable and cited more frequently.
  • Expansion of research network: Special Issues facilitate connections among authors, fostering scientific collaborations.
  • External promotion: Articles in Special Issues are often promoted through the journal's social media, increasing their visibility.
  • e-Book format: Special Issues with more than 10 articles can be published as dedicated e-books, ensuring wide and rapid dissemination.

Further information on MDPI's Special Issue polices can be found here.

Published Papers (2 papers)

Order results
Result details
Select all
Export citation of selected articles as:

Research

17 pages, 3460 KiB  
Article
Reservoir Computing Using Measurement-Controlled Quantum Dynamics
by A. H. Abbas and Ivan S. Maksymov
Electronics 2024, 13(6), 1164; https://doi.org/10.3390/electronics13061164 - 21 Mar 2024
Cited by 2 | Viewed by 1701
Abstract
Physical reservoir computing (RC) is a machine learning algorithm that employs the dynamics of a physical system to forecast highly nonlinear and chaotic phenomena. In this paper, we introduce a quantum RC system that employs the dynamics of a probed atom in a [...] Read more.
Physical reservoir computing (RC) is a machine learning algorithm that employs the dynamics of a physical system to forecast highly nonlinear and chaotic phenomena. In this paper, we introduce a quantum RC system that employs the dynamics of a probed atom in a cavity. The atom experiences coherent driving at a particular rate, leading to a measurement-controlled quantum evolution. The proposed quantum reservoir can make fast and reliable forecasts using a small number of artificial neurons compared with the traditional RC algorithm. We theoretically validate the operation of the reservoir, demonstrating its potential to be used in error-tolerant applications, where approximate computing approaches may be used to make feasible forecasts in conditions of limited computational and energy resources. Full article
Show Figures

Figure 1

15 pages, 1777 KiB  
Article
Priority-Aware Resource Management for Adaptive Service Function Chaining in Real-Time Intelligent IoT Services
by Prohim Tam, Sa Math and Seokhoon Kim
Electronics 2022, 11(19), 2976; https://doi.org/10.3390/electronics11192976 - 20 Sep 2022
Cited by 7 | Viewed by 1897
Abstract
The growth of the Internet of Things (IoT) in various mission-critical applications generates service heterogeneity with different priority labels. A set of virtual network function (VNF) orders represents service function chaining (SFC) for a particular service to robustly execute in a network function [...] Read more.
The growth of the Internet of Things (IoT) in various mission-critical applications generates service heterogeneity with different priority labels. A set of virtual network function (VNF) orders represents service function chaining (SFC) for a particular service to robustly execute in a network function virtualization (NFV)-enabled environment. In IoT networks, the configuration of adaptive SFC has emerged to ensure optimality and elasticity of resource expenditure. In this paper, priority-aware resource management for adaptive SFC is provided by modeling the configuration of real-time IoT service requests. The problem models of the primary features that impact the optimization of configuration times and resource utilization are studied. The proposed approaches query the promising embedded deep reinforcement learning engine in the management layer (e.g., orchestrator) to observe the state features of VNFs, apply the action on instantiating and modifying new/created VNFs, and evaluate the average transmission delays for end-to-end IoT services. In the embedded SFC procedures, the agent formulates the function approximator for scoring the existing chain performance metrics. The testbed simulation was conducted in SDN/NFV topologies and captured the average of rewards, delays, delivery ratio, and throughput as −48.6666, 10.9766 ms, 99.9221%, and 615.8441 Mbps, which outperformed other reference approaches, following parameter configuration in this environment. Full article
Show Figures

Figure 1

Back to TopTop