Data Transmission Reduction Model for cloud-based IoT Systems

Elouali, Aya; Mora, Higinio; Gimeno, Francisco José Mora

doi:10.1109/smartiot52359.2021.00046

Cited by 8 publications

(5 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Accordingly, the goal is to maximize the accuracy of the health assessment process while minimizing the amount of information transmitted over the monitoring network. The transmission of data toward the aggregation unit impacts, in fact, the network's energy consumption [4]. Placing the aggregation unit far from the sensors implies power-hungry long-range communications, encouraging configurations where the centralizing node is as close as possible to the sensing devices.…”

Section: Introductionmentioning

confidence: 99%

Combining Compressed Sensing and Neural Architecture Search for Sensor-Near Vibration Diagnostics

Ragusa,

Zonzini,

Gastaldo

et al. 2024

IEEE Trans. Ind. Inf.

View full text Add to dashboard Cite

Compressed sensing (CS) for sensor-near vibration diagnostics represents a suitable approach for the design of network-efficient structural health monitoring systems. This article presents a solution for vibration analysis based on deep neural networks (DNNs) trained on compressed data. The envisioned maintenance system consists of a network of sensing nodes orchestrated by a very constrained centralizing unit. The latter is equipped with a microcontroller unit (MCU) that predicts the health state using the aggregated information. As a major contribution, the DNN architectures are generated automatically from the data through a procedure inspired by hardware-aware (HW) neural architecture search (NAS), called as HW-NAS-CS, which is uniquely refined with additional constraints that consider both the peculiarities of CS parameters and the limitation of embedded devices. The proposed approach has been validated using two real-world SHM datasets for vibration damage identification and eventually deployed on a low-end computing platform (the STM32L5 MCU). Results demonstrate that DNNs combined with adapted CS schemes can attain classification scores always above 90% even in case of very huge compression levels (higher than 64x): these performances significantly improve the ones attained by state-of-the-art approaches in the field, with the utmost advantage of being portable on embedded devices.

show abstract

Section: Introductionmentioning

confidence: 99%

Combining Compressed Sensing and Neural Architecture Search for Sensor-Near Vibration Diagnostics

Ragusa,

Zonzini,

Gastaldo

et al. 2024

IEEE Trans. Ind. Inf.

View full text Add to dashboard Cite

show abstract

“…Notably, its application spans several sectors like in the monitoring and control of machinery through advanced technologies such as digital twins [4], smart factories, and cities [5,6], healthcare [7,8], agriculture [9,10] and in many other fields. IoT's success in all the above application areas, however, depends heavily on the ability to connect with many devices, collect, store, and process their data, and respond to changing conditions in real-time and this requires significant processing power and computational resources [11][12][13] In a typical cloud architecture for IoT data processing, raw data emanating from end-user devices are transmitted through high bandwidth data networks to remote servers for processing, storage, knowledge extraction, decision making, and finally, sending results to the same or other nodes for actuation [14,15]. This process flow however, introduces a heavy latency overhead making the inability to meet the latency, turnaround time and reliability demands of user applications a foremost challenge of IoT solutions especially in timecritical and mission-critical IoT scenarios that demand data processing within defined time constraints and reliability thresholds [16][17][18].…”

Section: Introductionmentioning

confidence: 99%

A Real-Time Task Balancing Strategy for IoT Networks Using Ant Colony Optimization

Wilson,

Nunoo-Mensah,

Boateng

et al. 2024

Preprint

View full text Add to dashboard Cite

IoT devices relying on cloud-based computing resources for data processing are causing unsustainable growth and latency, thus making their dependence on these computing resources a foremost challenge to real-time and time-critical IoT applications. Researchers have proffered several approaches, but notable amongst them are the bio-inspired optimisation methods due to their ability to model stochastic environments efficiently. This paper proposes a bio-inspired load-balancing optimisation approach for reducing latency in IoT networks. The paper leverages the concept of “stigmergy” in social ants to create a metric for assessing the computational capacity, reliability, and availability of a node’s computing resources in a decentralised Internet of Things (IoT) network. The main contribution of this work is the design of a decentralised serverless computing paradigm for stateful(reliable) task offloading and data processing among end-user nodes in an IoT environment. A novel bio-inspired algorithm is designed for latency reduction and optimal load-balancing in any collaborative data processing environment without a centralised server. This work also introduces the concept of computing pheromones as a metric for assessing computation resource capacity and availability in decentralised data processing environments, which paves the way for more efficient and reliable IoT data processing solutions. Experimental results show the effectiveness of the proposed IoT data processing approach, highlighting improvements in response time and turnaround time compared to existing approaches.

show abstract

“…We also define generalized evaluation metrics to evaluate each projection. This paper is a more in-depth study of our previous paper [ 11 ] where we briefly explained the general idea of this model and it is organized as follows: in " Section 2: related work " section we review most relevant works in terms of proposing solutions to improve the functioning of IoT systems and we determine how our work is situated compared to them. Section 3: data transmission reduction model section presents the proposed solution, its components as well as generalized evaluation formulas.…”

Section: Introductionmentioning

confidence: 99%

“…In our previous paper [ 11 ], we described the problematic and the general idea of the proposed solution as well as evaluation metrics that permit to calculate the effectiveness of the solution. In this paper, we describe the proposed method more in detail.…”

Section: Introductionmentioning

confidence: 99%

Data transmission reduction formalization for cloud offloading-based IoT systems

2023

Self Cite

View full text Add to dashboard Cite

Computation offloading is the solution for IoT devices of limited resources and high-cost processing requirements. However, the network related issues such as latency and bandwidth consumption need to be considered. Data transmission reduction is one of the solutions aiming to solve network related problems by reducing the amount of data transmitted. In this paper, we propose a generalized formal data transmission reduction model independent of the system and the data type. This formalization is based on two main ideas: 1) Not sending data until a significant change occurs, 2) Sending a lighter size entity permitting the cloud to deduct the data captured by the IoT device without actually receiving it. This paper includes the mathematical representation of the model, general evaluation metrics formulas as well as detailed projections on real world use cases.

show abstract

Data Transmission Reduction Model for cloud-based IoT Systems

Cited by 8 publications

References 17 publications

Combining Compressed Sensing and Neural Architecture Search for Sensor-Near Vibration Diagnostics

Combining Compressed Sensing and Neural Architecture Search for Sensor-Near Vibration Diagnostics

A Real-Time Task Balancing Strategy for IoT Networks Using Ant Colony Optimization

Data transmission reduction formalization for cloud offloading-based IoT systems

Contact Info

Product

Resources

About