Edge Server Allocation Through Composite Connection Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for allocating edge servers to user terminals in cloud applications, such as cloud gaming, often result in imbalanced server usage, with some servers becoming overloaded while others remain idle, due to considerations based on single dimensions like proximity or network quality, leading to suboptimal user experiences.
Innovation Solution
A method and system that calculates a connection reference score for multiple edge servers based on various dimensions like network quality, resource availability, and operation cost, ensuring only suitable servers are included in a candidate set from which the terminal selects the best one for optimal service provision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If edge servers are allocated based on single dimension (proximity or network quality), then allocation simplicity is maintained, but server usage becomes imbalanced with some servers overloaded and others idle
Solution Approach 1:
The patent changes the allocation parameter from single-dimension (proximity or network quality only) to multi-dimension by introducing a comprehensive scoring mechanism that evaluates multiple parameters including network quality, resource availability, and operation cost. This allows the system to select edge servers based on overall suitability rather than单一 criterion, resolving the contradiction between allocation simplicity and server usage balance
Solution Approach 2:
The patent creates a composite evaluation index by combining multiple independent evaluation dimensions (network quality, resource availability, operation cost) into a unified scoring system. This composite approach enables balanced server allocation by considering multiple factors simultaneously, preventing both overload and idle situations while maintaining systematic allocation
2Reliability
If multiple edge servers are deployed to provide stable network quality, then network stability is improved, but the complexity of server selection and allocation increases
Solution Approach 1:
The patent implements self-service by enabling the terminal to autonomously evaluate and score multiple edge servers based on predefined criteria (network quality, resource availability, operation cost). The terminal independently calculates comprehensive scores and selects the most suitable server without requiring complex centralized coordination, thus maintaining network stability while reducing selection complexity
Solution Approach 2:
The patent applies preliminary action by pre-establishing evaluation criteria and scoring rules for edge server selection. The system prepares the evaluation framework in advance, allowing terminals to quickly and consistently assess servers using predetermined metrics, thereby simplifying the selection process while ensuring reliable and stable server choices
3Speed
If the closest edge server is selected for terminal connection, then network delay is reduced, but server load becomes unbalanced with closest servers overloaded
Solution Approach 1:
The patent changes the selection parameter from purely proximity-based to a comprehensive scoring system that includes proximity (network delay) as one factor among multiple others (resource availability, operation cost). This allows terminals to select servers that offer optimal overall performance rather than simply the closest server, distributing load more evenly while maintaining acceptable network delay
Solution Approach 2:
The patent adds new evaluation dimensions beyond physical proximity by incorporating resource availability and operation cost into the selection criteria. This multi-dimensional approach enables the system to select servers based on comprehensive suitability rather than single-dimension proximity, balancing network delay requirements with server load distribution across the network
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
Disclosed in the embodiments of the present application are a service processing method and apparatus, and a storage medium. The method comprises: upon receiving an allocation request sent by a target terminal, taking each edge server as an edge server to be determined, and acquiring N pieces of selection reference information, that is related to the target terminal, of said edge server and running reference information that is required for running a target cloud application; if it is determined on the basis of the N pieces of selection reference information and the running reference information that said edge server satisfies the requirement for providing the target terminal with a running service for the target cloud application, determining a connection reference score of said edge server on the basis of the N pieces of selection reference information and a connection reference score determination rule; storing, in a candidate set, the connection reference score and identification information of said edge server; and sending the candidate set to the target terminal, such that the target terminal selects a target edge server from the candidate set; and the target edge server providing the target terminal with the running service of the target cloud application. Thus, connectable edge servers are uniformly allocated to the terminal.