Learning-Assisted Write Latency Optimization for Mobile Storage

I/O activities of mobile storage are highly synchronous. Flash garbage collection activities in mobile storage introduce extra delay to write requests and negatively impact on user perceived-latency. Runtime write demand is subject to correlation between multiple parameters, such as network connecti...

Full description

Saved in:
Bibliographic Details
Main Authors: Tsai, Wei-Chu, Wu, Sung-Ming, Chang, Li-Pin
Format: Conference Proceeding
Language:English
Subjects:
Correlation
Global Positioning System
Mobile storage
NAND flash
Optimization
Prediction algorithms
Smart phones
smartphones
Training
Wireless fidelity
XG-Boost
Online Access:Request full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:I/O activities of mobile storage are highly synchronous. Flash garbage collection activities in mobile storage introduce extra delay to write requests and negatively impact on user perceived-latency. Runtime write demand is subject to correlation between multiple parameters, such as network connectivity, GPS coordinates, and current time. We propose predicting write demand with a learning algorithm, XGBoost, and conducting background, rate-based garbage collection to optimize write latency without premature, excessive flash erasure. Our method reduced the 99-th percentile write latency by 56% compared to on-demand garbage collection and decreased flash erase count by 51% compared to unconditional background garbage collection.
ISSN:2325-1301
DOI:10.1109/RTCSA.2019.8864577