An efficient mapping technique for flash memory using grouped blocks

Abstract

Recently, Flash memory is adopted by many devices as it has very useful features. However, it has a critical drawback of “erase-before-write” feature. This problem incurs the unexpected performance degradation. To solve the upper serious disadvantage of flash memory, we require the software layer known as FTL (Flash Translation Layer) and the RAM (Random Access Memory). In this paper, we propose an economic mapping technique that the size of the mapping table is shrunk by a specific calculation. Grouped Block reduces the cost of flash memory and improves the performance by using the optimized mapping table, delaying the merge operation and reducing the number of erase operation. The results of experiments show that the size of the group mapping is more economical than that of block mapping and it demonstrates more improved performance compared to FAST.