A Novel Data Aggregation Preprocessing Algorithm in Flash Memory for Iot Based Power Grid Storage System

Abstract

NAND flash memory has the advantages of strong shock resistance, low power consumption, non-volatility, and high performance, and it is gradually applied to embedded systems and enterprise servers, such as the IoT-based power grid storage system. Therefore, research on the flash translation layer strategy has become a popular trend for solid-state storage devices. The mapping granularity of different FTL policies is different in the methods of garbage collection. Furthermore, there is a problem of “write amplification” in the write operation of the flash memory device. This paper proposes a data aggregation preprocessing algorithm that can aggregate as many active pages as possible and reduce the proportion of write amplification. The algorithm is implemented and tested on the flash simulation platform FlashSim. The results indicate that our algorithm can effectively improve the performance of the IoT-based power grid storage system, reducing redundant write operations, the number of physical block erasures, the number of physical page read and write times, and the system response time, and extending the life of solid-state devices.