A Novel Computation Offloading Under 6G LEO Satellite-UAV-based IoT

• Published in: 2022 3rd International Conference on Computation, Automation and Knowledge Management (ICCAKM) pp. 1 - 6
• Main Authors: Jaiswal, Kriti, Dahiya, Anshul, Saxena, Surabhi, Singh, Virendra, Singh, Archita, Kushwaha, Arun
• Format: Conference Proceeding
• Language: English
• Published: IEEE 15.11.2022
• Subjects: Autonomous aerial vehicles; Cloud computing; Computation Offloading; Data processing; Deep learning; Internet of Things; Internet of Things (IoT); Low Earth Orbit (LEO); Low earth orbit satellites; Satellite; Satellites; Unmanned Aerial Vehicle (UAV)
• DOI: 10.1109/ICCAKM54721.2022.9990352

The Internet of Things (IoT) has become a fundamental and omnipresent component of modern life, such as the infrastructure of intelligent transport systems and smart cities. All the next-generation IoT systems necessitate cognitive data processing that could be accomplished promptly through satellite communication. The edge server performs data processing in satellite communication, which is predicted to minimize service delay. However, in prevailing satellite technology, substantial transmission latency, significant route loss, and the resource, as well as energy limitations of IoT devices, all pose challenges to the rigorous service demands for latency and throughput in the upcoming 6G age. To tackle these challenges, 6G IoT will rely heavily on technologies such as space-air-ground integrated networks (SAGINs), edge computing, deep learning, machine learning, blockchain, and energy harvesting. Combining communication technology with Unmanned Aerial Vehicle (UAV) has recently been identified as one of the potential strategies in the future network. UAVs may provide a variety of services; however, the UAV's onboard computational resources and battery backup are constrained. The Researchers in this article examine satellites as well as UAVs to deliver wireless IoT device cloud computing and edge computing services, respectively. Grounded IoT devices can conduct operations locally or offload them to UAV-based edge servers as well as distant cloud servers via satellites. The major emphasis is on the computational offloading dilemma and considering deep learning methods to optimize the task success rate considering energy dynamics and channel circumstances. The suggested AI-based offloading approach outperformed conventional methods, indicating that it can determine the most energy-efficient offloading policy.