Trusted Metering for Cloud Workloads at Edge Devices
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Accurate and secure workload metering of cloud-provided workloads executing on edge computing devices is challenging due to the potential for users to circumvent metering efforts, leading to gross approximations in billing and resource management.
Innovation Solution
Implementing a trusted metering application within a trusted execution environment on edge computing devices to monitor and manage execution metrics of application modules, ensuring secure and accurate metering by enforcing dependencies between the application and the metering application.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If cloud-provided workloads are executed on edge computing devices, then workload execution flexibility and distributed computing capability are improved, but accurate and secure workload metering becomes difficult
Solution Approach 1:
A trusted metering application is introduced as an intermediary component that runs within the untrusted edge computing device environment but maintains secure measurement capabilities. This trusted application acts as a mediator between the workload execution and the cloud provider, accurately measuring resource consumption without requiring the edge device to be trusted. The trusted metering application isolates the measurement function from the untrusted execution environment, enabling precise metering while maintaining workload execution flexibility.
2Ease of operation
If traditional cloud metering approaches are used on edge devices, then implementation simplicity is maintained, but users can circumvent metering efforts leading to billing inaccuracies
Solution Approach 1:
The system segments the metering functionality into a separate trusted metering application that is distinct from both the cloud provider infrastructure and the edge device operating system. This segmentation isolates the critical measurement function within its own protected context, preventing users from circumventing metering by modifying other system components. The trusted metering application maintains secure hooks into resource allocation while remaining implementationally straightforward from the cloud provider perspective.
3Measurement precision
If a trusted execution environment is implemented on edge devices, then secure and accurate workload metering is achieved, but system complexity increases
Solution Approach 1:
The trusted metering application serves as an intermediary layer that abstracts the complexity of secure measurement from both the cloud provider and the edge device user. Rather than requiring complex changes to the entire edge device architecture, the trusted application provides a focused secure measurement capability that interfaces with existing workload execution mechanisms. This intermediary approach achieves accurate metering while containing the complexity increase to a specific, manageable component.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system meters execution of an application module at an edge computing device. A secure workload package is transmitted securely from a workload provisioning service to the edge computing device. The secure workload package includes the application module, a trusted metering application, and a provisioning service authentication token. The provisioning service authentication token is verified in the secure workload package based on an edge device authentication token generated at the edge computing device. The trusted metering application is executed in a trusted execution environment of the edge computing device, responsive to verifying the provisioning service authentication token. The application module of the edge computing device is executed, wherein the trusted metering application is configured to monitor execution metrics of the application module on the edge computing device. The execution of the application module is managed based on the monitored execution metrics.