Edge Workload Scheduler Using Secure Telemetry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In edge computing networks, efficiently scheduling workloads across resource-constrained edge devices while ensuring low latency and meeting security and service level agreement requirements is challenging due to the complexity of managing telemetry data and resource utilization.
Innovation Solution
A telemetry-based scheduler that collects and analyzes telemetric data from edge devices, adjusts workloads, and computes adjustments for downstream devices to remediate performance deficiencies and ensure compliance with service level agreements, while verifying the authenticity of data to maintain security.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If workload scheduling is performed on resource-constrained edge devices, then low latency and deterministic responses are achieved, but device complexity and resource management complexity increase
Solution Approach 1:
A centralized scheduler component is introduced as an intermediary between multiple edge devices and the workload management system. This scheduler collects telemetry data from all edge devices, performs centralized analysis, and distributes scheduling decisions back to the appropriate devices. This approach enables complex scheduling logic to be implemented without increasing the computational burden on individual resource-constrained edge devices, thereby maintaining low latency responses while managing system complexity centrally.
2Adaptability or versatility
If multiple services with different security requirements utilize limited edge node resources simultaneously, then service versatility is improved, but security management complexity and resource allocation difficulty increase
Solution Approach 1:
The system implements quality-of-service (QoS) policies that assign different security levels and resource allocation rules to different services based on their specific requirements. Each service is tagged with security attributes and resource entitlements, allowing the scheduler to apply localized security policies and resource constraints specific to each service type. This enables multiple services with varying security needs to coexist on the same edge nodes without requiring uniform security management across all services.
3Productivity
If telemetry data is collected and analyzed in real-time for workload scheduling, then workload scheduling efficiency is improved, but use of energy and processing resources increases
Solution Approach 1:
The scheduler implements periodic telemetry data collection and analysis cycles rather than continuous real-time monitoring. Telemetry data is gathered at predetermined intervals, and scheduling decisions are made based on these periodic updates. This approach maintains adequate workload scheduling efficiency by capturing meaningful changes in system state while significantly reducing the continuous energy consumption associated with constant real-time data processing and analysis.
4Reliability
If secure telemetry verification is performed to maintain security, then security reliability is improved, but processing time and computational overhead increase
Solution Approach 1:
Security credentials and device identities are pre-verified and authenticated before devices join the edge network and before telemetry data collection begins. Trust relationships are established in advance through secure enrollment procedures, allowing the scheduler to verify device authenticity using pre-computed security tokens and signatures. This preliminary security verification eliminates the need for time-consuming authentication checks during runtime telemetry verification, thereby maintaining high security reliability while minimizing processing time overhead during actual workload scheduling operations.
Data Source
AI summary
Methods, apparatus, systems and articles of manufacture are disclosed to schedule workloads based on secure edge to device telemetry by calculating a difference between a first telemetric data received from a first hardware device and an operating parameter and computing an adjustment for a second hardware device based on the difference between the first telemetric data and the operating parameter.