Loading [a11y]/accessibility-menu.js
Dynamic Pervasive Compute Orchestration using Information Centric Network | IEEE Conference Publication | IEEE Xplore

Dynamic Pervasive Compute Orchestration using Information Centric Network


Abstract:

With the rapid growth in number of connected devices, edge and fog computing paradigms become promising solutions for handling computation with low latency and reduced ba...Show More

Abstract:

With the rapid growth in number of connected devices, edge and fog computing paradigms become promising solutions for handling computation with low latency and reduced bandwidth usage. They allow to execute a computation across IoT devices, e.g., cameras, an edge aggregator/processor, and on the cloud. In this paper, we consider a pervasive computing scenario where end devices such as laptops, smart phones or roadside units are willing to offer compute services along with the traditional infrastructure-based edge computing resources. Our work uses an Information Centric Network (ICN), which is a promising technique for resource-constrained end devices to offload time-sensitive computing tasks to nearby distributed network resources, e.g., to an edge server. We expand the Named Data Networking (NDN), which is an ICN implementation, to operate in a dynamic resource-constrained multi-hop environment and propose i) a fast and efficient capability discovery scheme to assist each node in discovering other nodes’ capabilities in a distributed manner; and ii) efficient schemes for server/path selection, to optimize the end-to-end latency including discovery, data transmission and computation. Our results show that the proposed solutions can improve the total latency by around 50% for most of the tasks.
Date of Conference: 26-29 March 2023
Date Added to IEEE Xplore: 12 May 2023
ISBN Information:

ISSN Information:

Conference Location: Glasgow, United Kingdom

References

References is not available for this document.