Resource Allocation Strategies for Cloud-Edge-End Collaboration in Heterogeneous Edge Computing

With the continuous innovation of information technology, the Internet of Everything (IoE) is gradually being realized, and more IoT application scenarios have increasingly urgent requirements for high-performance computing and low-latency response. To address the bottleneck of insufficient computing power and energy constraints of terminal devices, edge computing can effectively enhance the task processing capability of terminal devices and improve service quality. However, existing edge computing network architectures, which often provide a single edge server computing capability based on terrestrial networks, cannot meet the task offloading requirements of terminal devices in ubiquitous network environments. This paper designs a cloud-edge-end collaboration space-air-ground integration multi-layer edge computing network architecture (CEEC-SAGIN), and based on this architecture, a heterogeneous network system of terminal devices, multi-layer edge computing services and cloud computing services are constructed. Then, a task processing model is constructed with the optimization objective of minimizing the task processing delay of the system, taking into account the task offloading ratio, computational resources and transmission resources involved in the task processing process of the terminal devices of the system. Further, a multi-layer task offloading and resource allocation algorithm (MTORA) is proposed to realize the optimal allocation of resources in system task processing. Finally, simulation experiments show that the CEEC-AGIN architecture proposed in this paper can meet the requirements of more applications, and the proposed model can improve the system task processing efficiency by up to roughly 15% and reduce the system task processing latency by up to approximately 10% compared with other resource allocation strategies.