Intelligent Monitoring of IoT Devices using Neural Networks

Chawla, Ashima; Babu, P Ravi; Gawande, Trushnesh; Aumayr, Erik; Jacob, Paul; Fallon, Sheila

doi:10.1109/icin51074.2021.9385543

Cited by 7 publications

(4 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It has become an essential option for IoT data analytics due to its reliability and stability in processing real-time streaming IoT data with minimum latency, and high throughput [5], [16]. Various works suggest its use in cloud native settings for scaling to real-time data streams coming from IoT devices [17]- [19], they also leverage compatible monitoring tools, such as Prometheus.…”

Section: A Iot Data Analytics Use Casesmentioning

confidence: 99%

A Comprehensive Performance Analysis of Stream Processing with Kafka in Cloud Native Deployments for IoT Use-cases

Pelle,

Szőke,

Fayad

et al. 2023

NOMS 2023-2023 IEEE/IFIP Network Operations and Management Symposium

View full text Add to dashboard Cite

The constant growth of the number of Internet of Things devices drives a huge increase in data that needs to be analyzed, at times in real time. Multiple platforms are available for delivering such data to analytics engines that can perform various operations on the data with low processing latency. These platforms can find their home in cloud native environments where high availability and scaling to the actual workload can be easily achieved. While the deployment environment is elastic, clusters still need to be adequately dimensioned to accommodate the components of the platforms even under high load.In this paper, we provide an analysis in this regard: we discuss key performance indicators of the popular Kafka message bus and the related Kafka Streams processing engine. Namely, we analyze latency, throughput, CPU and memory resource footprint aspects of these services under varying load and processing tasks that appear in Internet of Things applications. We find subsecond processing latency and linear but heavily task-dependent scaling behavior in the other performance indicators' case.

show abstract

Section: A Iot Data Analytics Use Casesmentioning

confidence: 99%

A Comprehensive Performance Analysis of Stream Processing with Kafka in Cloud Native Deployments for IoT Use-cases

Pelle,

Szőke,

Fayad

et al. 2023

NOMS 2023-2023 IEEE/IFIP Network Operations and Management Symposium

View full text Add to dashboard Cite

show abstract

“…The MSE is defined in (6), in which N is the number of data points, yi is the ith target value, and i y ˆ is the ith predicted value. Ideally, the output value of the network approaches the target output value as training progresses [21]. The MSE for the test set in each epoch is used to obtain the best prediction model (i.e., the model with the best generalization ability) [22].…”

Section: Neural Network (Nn)mentioning

confidence: 99%

Front-End Gateway System With Serial Communication Protocol Conversion and Edge Computing Platforms

Sung,

Tung,

Huang

et al. 2023

IEEE Access

View full text Add to dashboard Cite

This study presents a front-end gateway system with asynchronous and synchronous serial communication protocol conversion between the universal asynchronous receiver/transmitter (UART) protocol and inter-integrated circuit (I 2 C) protocol. The valid data and sensing data can be integrated using the I 2 C protocol and sent to an edge computing platform for automated data analysis. Our system reduces data processing time and the number of buses. The system has an edge computing platform that handles simple linear regression, base conversion, a neural network, and a text database and communicates with multiple peripheral devices. The I 2 C master and slave are constructed on the edge computing platform and implement arbitration by using the carrier-sense multiple access with collision avoidance protocol to prevent data collision. According to the results obtained from the Signal Tap logic analyzer in experiments conducted using a field-programmable gate array board, a completed 330-bit UART packet requires 755.2 s to be received, and the throughput is 436.97 kbps. By contrast, a 90-bit I 2 C packet requires 184.6 s to be received, and the throughput is 487.54 kbps. The front-end gateway sends integrated packets by using the I 2 C protocol, and the operating freuency (serial clock) of the I 2 C slave can reach up to 3.6 MHz bidirectionally. An integrated 153-bit packet requires 42.96 s to be received by the edge computing platform, and the throughput is 3.5614 Mbps, which is approximately 8.15 times higher than that of the UART packet. We also fabricated a front-end gateway ASIC by using the TSMC 90-nm 1P9M CMOS process.INDEX TERMS Serial communication protocol, universal asynchronous receiver/transmitter (UART), inter-integrated circuit (I 2 C), ZigBee, front-end gateway, edge computing platform, IoT network, application-specific integrated circuit (ASIC).

show abstract

“…At the first stage, a preliminary calculation of the neurons of the hidden layer takes place, the code for the Hidden_Layer_Calculation procedure is shown in Figure A1 (a) (see Appendix, lines [7][8][9][10][11][12][13][14][15][16][17][18][19]. The procedure calculates the weight coefficients W1 using Equaitons (3,4), then the arrays Y and W1 are multiplied in accordance with Equation (1).…”

Section: Implementation Of the Lognnet Network On The Arduino Uno Boardmentioning

confidence: 99%

“…The integration of IoT and artificial intelligence (AI) will allow for intelligent pre-processing of data on the peripheral devices without sending information to the cloud, reflecting the key area of the contemporary scientific research in the field of the Internet of things [8] and edge computing [9]. The significant research task is the development of new neural network technologies that could be successfully implemented in the IoT [10], widely used in daily activities, in healthcare [11], in designing robots [12] and in a smart homes [13], while consuming less computing resources and RAM memory.…”

Section: Introductionmentioning

confidence: 99%

Recognition of handwritten MNIST digits on low-memory 2 Kb RAM Arduino board using LogNNet reservoir neural network

Izotov

Velichko

Ivshin³

et al. 2021

IOP Conf. Ser.: Mater. Sci. Eng.

View full text Add to dashboard Cite

The presented compact algorithm for recognizing handwritten digits of the MNIST database, created on the LogNNet reservoir neural network, reaches the recognition accuracy of 82%. The algorithm was tested on a low-memory Arduino board with 2 Kb static RAM low-power microcontroller. The dependences of the accuracy and time of image recognition on the number of neurons in the reservoir have been investigated. The memory allocation demonstrates that the algorithm stores all the necessary information in RAM without using additional data storage, and operates with original images without preliminary processing. The simple structure of the algorithm, with appropriate training, can be adapted for wide practical application, for example, for creating mobile biosensors for early diagnosis of adverse events in medicine. The study results are important for the implementation of artificial intelligence on peripheral constrained IoT devices and for edge computing.

show abstract

Intelligent Monitoring of IoT Devices using Neural Networks

Cited by 7 publications

References 3 publications

A Comprehensive Performance Analysis of Stream Processing with Kafka in Cloud Native Deployments for IoT Use-cases

A Comprehensive Performance Analysis of Stream Processing with Kafka in Cloud Native Deployments for IoT Use-cases

Front-End Gateway System With Serial Communication Protocol Conversion and Edge Computing Platforms

Recognition of handwritten MNIST digits on low-memory 2 Kb RAM Arduino board using LogNNet reservoir neural network

Contact Info

Product

Resources

About