RSS-Based Memory Monitoring for Low-Overhead PSS and USS Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing memory monitoring methods incur high resource overheads, leading to issues such as freezing and overheating in terminal devices due to frequent polling of proportional set size (PSS) or unique set size (USS) of processes.
Innovation Solution
Obtain the resident set size (RSS) directly and use a preset correction value to determine PSS or USS, reducing the need for direct polling and minimizing resource overheads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If PSS or USS is polled frequently to obtain real-time memory monitoring results, then measurement precision is improved, but resource overhead increases causing device freezing and overheating
Solution Approach 1:
The patent pre-calculates and stores the difference values between PSS/RSS and USS/RSS relationships in advance. During monitoring, instead of calculating PSS or USS from scratch through resource-intensive operations, the system directly retrieves these pre-computed difference values and applies them to the currently obtained RSS to quickly derive PSS or USS estimates, thus reducing real-time computational overhead while maintaining measurement accuracy.
Solution Approach 2:
The patent creates a simplified model by copying the essential relationship patterns between memory metrics (PSS-RSS and USS-RSS differences) without replicating the full complexity of original memory calculation mechanisms. These copied difference values are stored in a lookup structure, allowing the system to replicate accurate memory measurement results through simple arithmetic operations rather than complex memory analysis, significantly reducing resource consumption during frequent monitoring operations.
2Measurement precision
If PSS or USS is polled frequently to obtain real-time memory monitoring results, then measurement precision is improved, but device stability deteriorates due to freezing and overheating
Solution Approach 1:
The patent pre-calculates and stores the difference values between PSS/RSS and USS/RSS relationships in advance. During monitoring, instead of calculating PSS or USS from scratch through resource-intensive operations, the system directly retrieves these pre-computed difference values and applies them to the currently obtained RSS to quickly derive PSS or USS estimates, thus reducing real-time computational overhead while maintaining measurement accuracy.
Solution Approach 2:
The patent creates a simplified model by copying the essential relationship patterns between memory metrics (PSS-RSS and USS-RSS differences) without replicating the full complexity of original memory calculation mechanisms. These copied difference values are stored in a lookup structure, allowing the system to replicate accurate memory measurement results through simple arithmetic operations rather than complex memory analysis, significantly reducing resource consumption during frequent monitoring operations.
3Device complexity
If RSS is obtained and PSS/USS is calculated directly without using correction values, then device complexity is reduced, but measurement precision decreases
Solution Approach 1:
The patent pre-calculates and stores the difference values between PSS/RSS and USS/RSS relationships in advance. During monitoring, instead of calculating PSS or USS from scratch through resource-intensive operations, the system directly retrieves these pre-computed difference values and applies them to the currently obtained RSS to quickly derive PSS or USS estimates, thus reducing real-time computational overhead while maintaining measurement accuracy.
Solution Approach 2:
The patent creates a simplified model by copying the essential relationship patterns between memory metrics (PSS-RSS and USS-RSS differences) without replicating the full complexity of original memory calculation mechanisms. These copied difference values are stored in a lookup structure, allowing the system to replicate accurate memory measurement results through simple arithmetic operations rather than complex memory analysis, significantly reducing resource consumption during frequent monitoring operations.
Data Source
Figure 1~3
Figure 4~5
Figure 6~8
AI summary
This application discloses a memory monitoring method and apparatus, and relates to the communication field. The memory monitoring method includes: obtaining a first resident set size RSS occupied by a process; and determining, based on the first RSS, first memory including a first PSS and/or a first USS. Resource overheads during monitoring by using the memory monitoring method are low, so that smooth running of a terminal device can be ensured.