Answers are grounded in this paper and cite their sources.
Deep Neural Networks (DNNs) are steadily gaining relevance in several safetyand time-critical applications in the domain of cyber-physical systems.However, their resource intensive nature poses significant challenges to their deployment in resource-constrained embedded settings.Orthogonally, with this increasing complexity of workloads, real-time resource management techniques must also consider functional dependencies among them.Focussing on such research questions, this special issue titled "Performance Guarantees in Dynamic Networks and Graph Algorithms", presents two papers which are extended versions of some of the outstanding papers presented at IEEE Real-Time Systems Symposium (RTSS) 2022.The IEEE RTSS is a premier conference in the field of real-time systems and is a venue for researchers and practitioners to showcase innovations covering all aspects of real-time systems.In 2022, RTSS celebrated its 43rd anniversary, and it continued the trend of making RTSS an expansive and inclusive event striving to embrace new and emerging areas of real-time systems research.RTSS 2022 received a total of 128 submissions, out of which 37 papers were accepted to appear at the conference, and a further selection of 4 outstanding papers from this shortlist was done by a best-paper committee.The authors of these outstanding papers were invited to submit an extension of their work to this special issue.The first paper in this special issue, titled "Inference Serving with End-to-End Latency SLOs over Dynamic Edge Networks", was written by Vinod Nigade, Pablo Bauszat, Henri Bal and Lin Wang.It delves into deep learning (DL) inference serving for mobile and IoT applications, essential for technologies like augmented reality and autonomous driving.DL models, such as DNNs, require intensive computations, and are challenging to deploy on devices with limited capabilities.
No takes yet. Share an insight, caveat, or question.
Easwaran et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: