适用于FPGA与Nand flash阵列星载固态存储器的坏块管理方法

    A novel bad block management methodology suitable for FPGA and Nand flash array based spaceborne solid-state memory

    • 摘要: 随着遥感、雷达、通信技术的快速发展,卫星有效载荷速率爆发式增长,星载固态存储器需要存储的数据量越来越大,存储速率越来越高,存储读写次数越来越多。作为当前星载固存的主流产品,基于FPGA和Nand flash的星载固态存储器所需的Nand flash数量越来越多,单片容量越来越大。受限于Nand flash的工艺特性,基于FPGA和Nand flash的固态存储器在其生命周期内会产生更多的坏块,从而影响记录载荷数据的正确性。针对星载固态存储器使用过程中产生再生坏块问题,提出了一种基于实时坏块检测、标注和坏块数据自主搬移、坏块自主回收的固态存储器坏块管理方案,有效地消除了再生坏块对载荷数据正确性的影响,降低了再生超级坏块数据的错误扩散和重复错误概率。方案已经过测试验证并已具有多个型号飞行经历,是一种简单、高效、可靠的星载固态存储器坏块管理方法。

       

      Abstract: With the rapid development of remote sensing, radar and communication technology, the satellite payload rate has exploded. Spaceborne solid-state memory needs to store more and more data, the storage rate is higher and higher, and the storage read and write times are more and more. As the current mainstream product of spaceborne solid-state memory, the solid-state memory based on FPGA and Nand flash requires more and more Nand flash, and the single-chip capacity is larger and larger. Limited by the workmanship characteristics of Nand flash, the solid-state memory based on FPGA and Nand flash will produce more bad blocks in its life cycle, which will affect the accuracy of the satellite recorded data. Aiming at the problem of regenerating bad blocks in the use of spaceborne solid-state memory, this paper proposes a bad block management scheme of solid-state memory based on real-time bad block detection, marking, bad block data autonomous moving and bad block autonomous recycling. This bad block management scheme effectively eliminates the influence of regenerating bad blocks on the correctness of load data, and reduces the error diffusion and repeated error probability of regenerating super bad block data. This scheme has been tested and verified, and has flight experience of several satellites. It is a simple, efficient and reliable method for bad block management of spaceborne solid-state memory.