MEC Subservice Scheduling With Distributed Protection Placement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile edge computing (MEC) systems lack effective methods for scheduling distributed dedicated protection services, which are crucial for resource-intensive and delay-sensitive applications, leading to server overloading and increased delays without considering service division and joint optimization of computing resources and protection resources.
Innovation Solution
A method involving a double round-robin scheduling policy is employed to select MEC servers for subservices and their protection subservices, using a shortest routing algorithm and integer linear programming optimization to minimize total service delay while ensuring resources are not wasted or overloaded.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If services are concentrated on single MEC servers, then service processing simplicity is improved, but server overloading and increased delays occur
Solution Approach 1:
The patent divides a service into multiple subservices and distributes them across different MEC servers. Each subservice can be processed independently by different servers, preventing any single server from becoming overloaded while maintaining efficient parallel processing, thus reducing overall service delay
Solution Approach 2:
The patent introduces a new dimension of service deployment by creating protection subservices that are deployed on different servers from their corresponding working subservices. This dimensional separation ensures that failure or overload on one server does not affect the entire service, improving both reliability and delay performance
2Reliability
If service division and distributed deployment are implemented, then service reliability is improved, but system complexity increases
Solution Approach 1:
The service is segmented into working subservices and protection subservices, with each type deployed on separate MEC servers. This segmentation ensures that if a server fails, the protection subservices can take over, improving reliability while keeping the segmentation logic manageable through clear separation of concerns
Solution Approach 2:
The patent changes the scheduling parameter from single-server assignment to dual-server assignment (working server + protection server). By systematically assigning subservices to different servers than their protection counterparts, the system achieves high reliability through a manageable parameter change rather than complex redundant systems
3Productivity
If protection subservices are deployed on the same server as working subservices, then resource utilization is improved, but service survivability decreases
Solution Approach 1:
The patent segments the server selection process into two distinct phases: first selecting a working server for the subservice, then selecting a different server for the protection subservice. This segmentation ensures physical separation of working and protection functions, guaranteeing service survivability even with limited server resources
Solution Approach 2:
Instead of deploying protection subservices on the same server to maximize resource utilization, the patent inverts the approach by deliberately deploying them on different servers. This inversion prioritizes service survivability over immediate resource utilization, ensuring that protection resources are available when needed
4Loss of time
If conventional distributed service scheduling is used, then transmission delay is minimized, but protection resource optimization is not considered
Solution Approach 1:
The patent merges two previously separate concerns - service scheduling for delay optimization and protection resource allocation for reliability - into a unified scheduling framework. The double round-robin policy simultaneously optimizes both working subservice placement and protection subservice placement, achieving both low delay and high reliability
Data Source
AI summary
The invention provides a method and system for scheduling a mobile edge computing-oriented distributed dedicated protection service. The method includes: dividing a service into subservices and generating corresponding protection subservices; and simultaneously selecting a working server for a subservice and selecting a MEC server for a protection subservice by using a double round-robin scheduling policy, where the server selected for the subservice is different from the server selected for the protection subservice. The system includes a large-scale network formed by MEC servers, where services and dedicated protection services corresponding to the services are arranged on the MEC servers by the above method. A total delay of a service is minimized on the premise of protecting the service, thereby avoiding a waste of computing resources or overloading, and implementing the division of mobile edge computing-oriented distributed services and the joint optimization of computing resources and protection computing resources of subservices.