Energy Efficient Neural Network Embedding in IoT over Passive Optical Networks

Alenazi, Mohammed M.; Yosuf, Barzan A.; El-Gorashi, Taisir E. H.; Elmirghani, Jaafar M. H.

doi:10.1109/icton51198.2020.9203403

Cited by 6 publications

(5 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This work builds on our earlier proposals in various areas such as distributed processing in the IoT/Fog [9]- [12], green core and DC networks [13]- [22] , [23]- [28], network virtualization and service embedding in core and IoT networks [29]- [32] and machine learning and network optimization for healthcare systems [33]- [36] and network coding in the core network [37], [38]. Our previous work in [39] dealt with the idea of generic service embedding in an IoT setting, we take the work further in this paper by refining the optimization model and abstracting the virtual service requests (VSRs) that comprise of multiple Virtual Machines (VMs) inter-connected in a virtual topology. This allowed us to scale up the MILP model and represent closer to realistic DNN inference workloads in the optimization framework.…”

Section: Introductionmentioning

confidence: 89%

Energy-Efficient AI over a Virtualized Cloud Fog Network

Yosuf

Mohamed

Alenazi

et al. 2021

Proceedings of the Twelfth ACM International Conference on Future Energy Systems

Self Cite

View full text Add to dashboard Cite

Deep Neural Networks (DNNs) have served as a catalyst in introducing a plethora of next-generation services in the era of Internet of Things (IoT), thanks to the availability of massive amounts of data collected by the objects on the edge. Currently, DNN models are used to deliver many Artificial Intelligence (AI) services that include image and natural language processing, speech recognition, and robotics. Accordingly, such services utilize various DNN models that make it computationally intensive for deployment on the edge devices alone. Thus, most AI models are offloaded to distant cloud data centers (CDCs), which tend to consolidate large amounts of computing and storage resources into one or more CDCs. Deploying services in the CDC will inevitably lead to excessive latencies and overall increase in power consumption. Instead, fog computing allows for cloud services to be extended to the edge of the network, which allows for data processing to be performed closer to the end-user device. However, different from cloud data centers, fog nodes have limited computational power and are highly distributed in the network. In this paper, using Mixed Integer Linear Programming (MILP), we formulate the placement of DNN inference models, which is abstracted as a network embedding problem in a Cloud Fog Network (CFN) architecture, where power savings are introduced through trade-offs between processing and networking. We study the performance of the CFN architecture by comparing the energy savings when compared to the baseline approach which is the CDC. CCS CONCEPTS• Computing methodologies → Machine learning; • Networks → Network architectures.

show abstract

Section: Introductionmentioning

confidence: 89%

Energy-Efficient AI over a Virtualized Cloud Fog Network

Yosuf

Mohamed

Alenazi

et al. 2021

Proceedings of the Twelfth ACM International Conference on Future Energy Systems

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, the study also highlighted the need for data security and privacy measures to be implemented, as the devices collected sensitive information such as occupancy patterns. The study by Alenazi et al proposes an energy-efficient neural network embedding technique in IoT over passive optical networks to enhance the performance of IoT-based applications while reducing energy consumption [18].…”

Section: Smart Homesmentioning

confidence: 99%

IoT and Energy

M. Alenazi

2024

Internet of Things - New Insights

View full text Add to dashboard Cite

The Internet of Things (IoT) has the potential to revolutionize energy management by enabling the collection and analysis of real-time data from various energy sources. This research paper investigates the impact of the Internet of Things (IoT) on energy management. The paper provides an overview of IoT and its potential applications in energy management, including improved efficiency, reduced costs, and better resource utilization. The benefits of using IoT for energy management and the major challenges that may arise in implementing IoT-enabled energy management are discussed. Potential solutions to these challenges, such as artificial intelligence and cloud computing, are presented, along with case studies of IoT-enabled energy management in different industries. The paper also analyzes the impact of IoT on energy efficiency in telecommunications and cloud infrastructure. Finally, the future outlook for IoT and energy management is discussed, including potential developments in edge computing, advanced analytics, and 5G networks. Overall, this paper highlights the potential of IoT to revolutionize energy management and provides insights into the challenges and opportunities of implementing IoT-enabled energy management solutions.

show abstract

“…In their work, they adopted mixed-integer linear programming (MILP) to optimize virtual machine placement in order to meet intensive demand and reduce power consumption. Alenazi et al [16] presented a framework of an energy-efficient neural network service for smart homes that was embedded in Internet of Things (IoT) devices over PONs. They utilized MILP to formulate the embedding problem that minimized the total power consumption of both networking and processing.…”

Section: Introductionmentioning

confidence: 99%

A Loading-Aware TDMA Sleep Scheme to Improve Power Consumption Performance on Medium to High Traffic for NG-EPON Networks

Liu

Wu²

2022

Sustainability

View full text Add to dashboard Cite

The inter-cycle sleep design previously proposed as a power-saving scheme for next-generation Ethernet passive optical networks (NG-EPONs) is able to effectively decrease the power consumption of heavy-load traffic. However, with a medium to high traffic load, these networks may still suffer from a high level of power consumption if no enhanced mechanisms are used. It was noted that the optical line terminal (OLT) often fully opens all communication channels at a medium to heavy load. Moreover, the number of opened channels was changed cyclically according to the traffic loading status. Accordingly, optical network units (ONUs) may be frequently allocated to different channels with changing loads. This may lead to inefficient operation and result in significant channel tuning delays. We thus propose a loading-aware time-division multiple access (TDMA) mechanism that allows the OLT to reserve a maximum bandwidth for each ONU when the network experiences heavy-load traffic. The performance results of our simulations reveal that the proposed scheme is able to reduce power consumption for targeted medium to high loads since ONUs under such loads can extend their sleep time because the cycle length is maximized. Moreover, the total delay is maintained at a relatively low level after applying the proposed scheme since the tuning delay is reduced significantly; however, the transmission delay is slightly increased due to the increased cycle length.

show abstract

Energy Efficient Neural Network Embedding in IoT over Passive Optical Networks

Cited by 6 publications

References 33 publications

Energy-Efficient AI over a Virtualized Cloud Fog Network

Energy-Efficient AI over a Virtualized Cloud Fog Network

IoT and Energy

A Loading-Aware TDMA Sleep Scheme to Improve Power Consumption Performance on Medium to High Traffic for NG-EPON Networks

Contact Info

Product

Resources

About