Rizvi and Chung attempt to use block access patterns to improve flash memory storage performance. In such file systems, the operating system divides each file into blocks, usually consisting of 4 kilobytes (KB). The flash memory operates at the level of 512-byte pages. The proposed system optimizes the allocation of pages to blocks so that active (hot) blocks require minimal updating of data structures on the drive, resulting in faster performance and lower wear on the physical file system.
Unfortunately, the writing in this paper is extremely difficult to follow. The authors struggle with the English language, and write very tersely. If you read this paper, prepare to spend a lot of time deciphering the text for the kernels of new ideas contained within.