SSD Random Write Arbitration Using LSTM Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing solid-state disk (SSD) technologies face challenges with low processing speed and poor consistency in random write operations, failing to meet current computing demands.
Innovation Solution
A method and apparatus that utilize a wear-leveling module to determine a target super logic unit, scan free and bad blocks, and iteratively update historically weighted data frames using a long short-term memory network, coupled with dynamic write arbitration to adjust read and write operations per second, improving input/output performance and consistency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional random write methods are used in SSD, then the implementation is simple, but the processing speed is low and data consistency is poor
Solution Approach 1:
The patent implements a feedback mechanism where the system continuously monitors free block quantities and adjusts write arbitration strategies dynamically. The wear-leveling module scans free blocks and bad blocks, updates historical weighted data frames, and adjusts write operations based on current SSD state, creating a closed-loop control system that adapts to changing conditions to maintain optimal performance
Solution Approach 2:
The patent transforms the static write arbitration mechanism into a dynamic one by introducing adjustable write arbitration that changes based on real-time SSD conditions. The system dynamically selects different arbitration strategies (round-robin, random, or sequential) based on the quantity of free blocks and historical performance data, allowing the system to adapt its behavior to current system state rather than using a fixed approach
2Reliability
If traditional random write methods are used in SSD, then the system complexity is low, but data consistency is poor
Solution Approach 1:
The feedback mechanism monitors free block quantities and adjusts write operations to maintain data consistency. By continuously updating historical weighted data frames and comparing current state with expected states, the system can detect and correct deviations that would lead to inconsistency, ensuring reliable data storage across different write scenarios
Solution Approach 2:
The patent performs preliminary wear-leveling operations before actual data writes by scanning and identifying suitable target blocks in advance. The wear-leveling module pre-processes the SSD structure by leveling remaining spaces and identifying optimal write locations, which prepares the system for subsequent writes and ensures consistent data placement patterns that improve overall data consistency
3Productivity
If wear-leveling and dynamic write arbitration are implemented, then input/output performance improves, but the computational complexity increases
Solution Approach 1:
The patent applies partial wear-leveling by focusing on leveling only the remaining spaces in super logic units rather than performing complete wear-leveling of the entire SSD. The wear-leveling module scans and levels portions of the storage structure that are most critical for performance, avoiding the computational overhead of exhaustive wear-leveling while still achieving significant performance improvements
Solution Approach 2:
The patent divides the SSD into super logic units and processes them individually through wear-leveling operations. By segmenting the large storage space into manageable super logic units, the system can perform wear-leveling and write arbitration on smaller, more manageable portions, reducing the computational complexity of each operation while maintaining overall effectiveness
Data Source
AI summary
A random write method includes: using a wear-leveling module to scan the number of free blocks and the number of bad blocks in a target super logic unit; using a lookup management module to iteratively update, according to the number of current remaining solid-state disk data frames, the number of historically weighted solid-state disk data frames in a long short-term memory network manner; using dynamic write arbitration to determine an adjustment stage based on the number of historically weighted solid-state disk data frames, and determining the expected number of read and write operations per second based on the adjustment stage; and re-updating the number of historically weighted solid-state disk data frames, and adjusting the actual number of read and write operations per second based on the re-updated number of historically weighted solid-state disk data frames and the expected number of read and write operations per second.