본문 바로가기
  • Home

Data Mining Based Data Pattern Analysis and Memory Pinning Scheme to Improve Response Time on Small Mobile Devices

  • Journal of Internet of Things and Convergence
  • Abbr : JKIOTS
  • 2025, 11(2), pp.151~156
  • Publisher : The Korea Internet of Things Society
  • Research Area : Engineering > Computer Science > Internet Information Processing
  • Received : March 18, 2025
  • Accepted : April 14, 2025
  • Published : April 30, 2025

Hyun-Seob Lee 1

1백석대학교

Accredited

ABSTRACT

In recent years, small mobile devices have been equipped with high-performance computing power and storage, and efficient data access is essential. In particular, flash memory provides high speeds for data processing, but it suffers from asymmetric read/write performance and repeated write/erase cycles that can lead to failure. Buffers have been used to solve these problems, but while write performance has improved, read performance still suffers from asymmetric response latency. To solve these problems, this paper proposes a method to keep frequently accessed data fixed in a buffer. This method analyzes data patterns based on data mining and allocates memory resources to maintain data in the analyzed patterns. In addition, we optimize the use of resources by keeping the fixed memory resources within 10% of the total storage capacity. Finally, we demonstrate through experiments that the performance and lifetime of storage devices are improved. The proposed data mining analysis technique and memory resource optimization technique are expected to contribute to the development of buffer management techniques optimized for small storage devices.

Citation status

* References for papers published after 2023 are currently being built.