Search for: All records

Federated learning is a novel paradigm allowing the training of a global machine-learning model on distributed devices. It shares model parameters instead of private raw data during the entire model training process. While federated learning enables machine learning processes to take place collaboratively on Internet of Things (IoT) devices, compared to data centers, IoT devices with limited resource budgets typically have less security protection and are more vulnerable to potential thermal stress. Current research on the evaluation of federated learning is mainly based on the simulation of multi-clients/processes on a single machine/device. However, there is a gap in understanding the performance of federated learning under thermal stress in real-world distributed low-power heterogeneous IoT devices. Our previous work was among the first to evaluate the performance of federated learning under thermal stress on real-world IoT-based distributed systems. In this paper, we extended our work to a larger scale of heterogeneous real-world IoT-based distributed systems to further evaluate the performance of federated learning under thermal stress. To the best of our knowledge, the presented work is among the first to evaluate the performance of federated learning under thermal stress on real-world heterogeneous IoT-based systems. We conducted comprehensive experiments using the MNIST dataset and various performance metrics, including training time, CPU and GPU utilization rate, temperature, and power consumption. We varied the proportion of clients under thermal stress in each group of experiments and systematically quantified the effectiveness and real-world impact of thermal stress on the low-end heterogeneous IoT-based federated learning system. We added 67% more training epochs and 50% more clients compared with our previous work. The experimental results demonstrate that thermal stress is still effective on IoT-based federated learning systems as the entire global model and device performance degrade when even a small ratio of IoT devices are being impacted. Experimental results have also shown that the more influenced client under thermal stress within the federated learning system (FLS) tends to have a more major impact on the performance of FLS under thermal stress. 

Chemical transformations near plasmonic metals have attracted increasing attention in the past few years. Specifically, reactions occurring within plasmonic nanojunctions that can be detected via surface and tip-enhanced Raman (SER and TER) scattering were the focus of numerous reports. In this context, even though the transition between localized and nonlocal (quantum) plasmons at nanojunctions is documented, its implications on plasmonic chemistry remain poorly understood. We explore the latter through AFM-TER-current measurements. We use two molecules: i) 4-mercaptobenzonitrile (MBN) that reports on the (non)local fields and ii) 4-nitrothiophenol (NTP) that features defined signatures of its neutral/anionic forms and dimer product, 4,4′-dimercaptoazobenzene (DMAB). The transition from classical to quantum plasmons is established through our optical measurements: It is marked by molecular charging and optical rectification. Simultaneously recorded force and current measurements support our assignments. In the case of NTP, we observe the parent and DMAB product beneath the probe in the classical regime. Further reducing the gap leads to the collapse of DMAB to form NTP anions. The process is reversible: Anions subsequently recombine into DMAB. Our results have significant implications for AFM-based TER measurements and their analysis, beyond the scope of this work. In effect, when precise control over the junction is not possible (e.g., in SER and ambient TER), both classical and quantum plasmons need to be considered in the analysis of plasmonic reactions