JP5343817B2 - Storage device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a memory device securing a data retention period on a data-by-data basis. <P>SOLUTION: A memory device 10 has: a plurality of blocks each having a function to store data, wherein stored data retention period is shortened according to the number of data rewrite times; and a wear leveling means 12 for performing wear leveling to level the number of data rewrite times for each of the plurality of blocks. The memory device 10 further has a block group dividing means 12 for dividing the memory means 10 into a plurality of block groups. The wear leveling means 12 performs wear leveling in a range of each divided block group. Preferably, the block group dividing means 12 allocates each block group for use on an application-by-application basis. <P>COPYRIGHT: (C)2011,JPO&amp;INPIT

Description

  The present invention relates to a storage device.

  A storage device such as a flash memory or an SD memory card can rewrite data, and the stored data is retained without being erased even after the power is turned off. However, in such a storage device, if writing / erasing is performed a certain number of times or more with respect to each block, a write error occurs, and thereafter, it becomes a bad block and cannot be used. For example, in the case of a NAND flash memory, a write error occurs when the number of times of writing / erasing to each block is 100,000 times for SCL and 10,000 times for MCL. If write errors occur frequently, the writable area that can be used decreases, and the required write size cannot be maintained. Therefore, it is necessary to limit the number of times of writing / erasing. Further, in such a storage device, when the number of data rewrites increases, the period during which data can be held is shortened (data retention is reduced). For example, in the case of a NAND flash memory, as shown in FIG. 5, when the number of write / erase times exceeds a predetermined number, the data retention decreases linearly as the number of write / erase times increases.

  Therefore, a certain number of write thresholds is set for each block, and when the number of writes of a block exceeds the number of thresholds, a process is performed to write to another block, and wear leveling is performed to equalize the number of write / erase times to each block. Is going. Patent Document 1 discloses wear leveling that distributes the rewriting frequency of all blocks in order to prevent rewriting from concentrating on a specific area on a NAND flash memory.

JP 2002-32256 A JP 2008-112285 A JP 2003-208352 A

  Data stored in the storage device may include data having a high rewrite frequency (data mainly written) and data having a low rewrite frequency (data mainly read). For example, in the case of a vehicle, there are data that is mainly read in such as map data and data that periodically rewrites a small amount of data such as vehicle diagnosis data. In the case of vehicle diagnosis data, for example, it is necessary to write 356 bytes in one record every 500 milliseconds, and to store data for a maximum of 5 days. Only 72,000 times of writing occurs. On the other hand, in the case of map data, the period during which data must be retained is long (yearly), and it is necessary to guarantee the data retention period or longer. When wear leveling is performed when data with a high rewrite frequency and data with a low rewrite frequency are mixed, the block in which the data with the low rewrite frequency is stored is the target block for rewriting the data with the high rewrite frequency. It may become. In this case, data with a low rewrite frequency is stored in a block in which data retention has deteriorated, and there is a possibility that the data cannot be held for a holding period necessary for the data.

  Therefore, an object of the present invention is to provide a storage device that can secure a necessary data holding period for each data.

A storage device according to the present invention has a function of storing data, and a plurality of blocks in which a period in which stored data can be held according to the number of data rewrites is shortened and the number of data rewrites for each of the plurality of blocks are leveled. Used storage area for calculating wear area leveling means for calculating a used area size for a period in which application data must be held for each application that stores data in the storage apparatus And a block group dividing unit that divides the storage device into a plurality of block groups. The block group dividing unit allocates a block group corresponding to the used area size calculated by the used area size calculating unit for each application, and wear leveling is performed. Means within the block group divided by the block group dividing means. And performing A leveling.

  This storage device has a plurality of blocks. Each block has a shorter period during which data can be held in accordance with the number of data rewrites. In the storage device, the data is divided into a plurality of block groups by the block group dividing means. Thereby, it can be divided into a plurality of block groups such as a block group for storing data with a high rewrite frequency and a block group for storing data with a low rewrite frequency. Incidentally, since each block group is composed of one or more blocks, the number of block groups <the number of blocks. In the storage device, the wear leveling means equalizes the number of data rewrites of the blocks included in the block group on a block group basis. As a result, wear leveling is performed separately for each block group such as a block group for storing data with a high rewrite frequency and a block group for storing data with a low rewrite frequency. As a result, data with a low rewrite frequency is not stored in a block in which data with a high rewrite frequency is stored. As described above, in the storage device, it is possible to secure a necessary data holding period for each data by dividing into a plurality of block groups and performing wear leveling in units of block groups.

  Rewrite frequency, data retention period, etc. differ for each application. Therefore, in this storage device, a block group to be used for each application is assigned by the block group dividing unit, and the data rewrite frequency of the blocks included in the block group is leveled by the wear leveling unit for each block group of each application. As described above, in the storage device, a data holding period can be secured for each application by dividing the application into a plurality of block groups and performing wear leveling for each application block group.

  In the storage device of the present invention, the storage device is preferably a flash memory. In the storage device of the present invention, the wear leveling means manages the data rewrite frequency of each block for each block group, and selects a block to be rewritten according to the data rewrite frequency.

  According to the present invention, a necessary data holding period can be ensured for each data by dividing into a plurality of block groups and performing wear leveling in units of block groups.

It is a block diagram of the storage device control apparatus which concerns on this Embodiment. It is explanatory drawing of the calculation method of the use area size of each service. It is an example of area | region allocation with respect to each service in the memory | storage device of FIG. It is an example of an internal structure of a NABD type flash memory. It is a graph which shows an example of the relationship between the write / erase frequency of NAND type flash memory, and data retention.

  Hereinafter, embodiments of a storage device according to the present invention will be described with reference to the drawings. In addition, the same code | symbol is attached | subjected about the element which is the same or it corresponds in each figure, and the overlapping description is abbreviate | omitted.

  In the present embodiment, the storage device according to the present invention is applied to a storage device control device mounted on a vehicle. The storage device control device according to the present embodiment includes a storage device (storage device), and controls writing / reading of service (application) data in each ECU [Electronic Control Unit] of the vehicle. Examples of services that store data in the storage device include a vehicle diagnosis service, a remote diagnosis service, and map data (road information).

  The storage device control apparatus 1 will be described with reference to FIGS. FIG. 1 is a configuration diagram of a storage device control apparatus according to the present embodiment. FIG. 2 is an explanatory diagram of a method for calculating the used area size of each service. FIG. 3 is an example of area allocation for each service in the storage device of FIG. FIG. 4 shows an example of the internal configuration of the NAND flash memory.

  For each service, the storage device control apparatus 1 calculates the use area size necessary for the durable years (period in which data must be retained) in consideration of the writing method of each service, and a block group corresponding to the use area size. (Area) is allocated, and wear leveling is performed in units of the allocated area. The storage device control device 1 includes a storage device 10, an IF [InterFace] unit 11, and an access control unit 12. In the present embodiment, each process of the access control unit 12 corresponds to a block group dividing unit and a wear leveling unit described in the claims. However, the block group dividing means may be performed by computer processing used in the design stage.

  The storage device 10 is a storage device provided in the vehicle, and is, for example, a flash memory (NAND type, NOR type) or an SD memory card. The storage device 10 is a non-volatile memory that can be rewritten and does not lose data even when the power is turned off. The storage device 10 has a large number of blocks. In each block, when the number of rewrites exceeds a certain number, the period during which data can be held becomes shorter as the number of rewrites increases. Note that the memory capacity of the storage device 10 is a memory capacity having a sufficient margin with respect to the total size of the used area sizes for all services for storing data.

  The IF unit 11 is an IF circuit of the storage device 10, and is, for example, an IF circuit corresponding to a NAND flash memory or an IF circuit corresponding to an SM memory card.

  The access control unit 12 is a file system control device (a device composed of CPU, RAM, software for performing each process of the file system, etc.), and controls writing / reading of data of each service to / from the storage device 10. In particular, the access control unit 12 performs used area size calculation processing, area allocation processing, and data writing processing.

  The use area size calculation process will be described. For each service that stores data in the storage device 10, the access control unit 12 considers the service writing method and calculates the use area size necessary for storing for the durable years. This calculation is performed in advance before starting the service. For services that are determined to be installed at the design stage, such as services such as vehicle diagnostic services and map data (road information), the use area size calculation processing is performed by a computer at the design stage, and the use area of each service is determined. The size may be calculated. In this case, the access control unit 12 performs the used area size calculation process only for a service added later by a distribution service or the like.

  As a specific calculation method, as shown in FIG. 2, the used area size is calculated based on the upper limit number of times of writing, the durable years, the physical configuration of the device (storage device 10), and the service writing method. The use area size is a size that allows data to be retained for the durable years and does not exceed the upper limit number of times when all blocks are written according to the service writing method.

  As the upper limit number of times of writing, an upper limit number of times at which data retention does not decrease is set based on the relationship between the number of times of writing / erasing of each device and data retention as shown in FIG. The durable years are basically set in consideration of a period that must be guaranteed as a service (in the case of a vehicle, generally about 7 to 10 years). In the case of a service that is determined to be updated in units of years, such as map data, the durable years are set with a margin from the updated years. By the way, although it is a durable life, if the period that must be guaranteed as a service is set on a daily or monthly basis, the period is set accordingly. The physical configuration of the device is set based on the specification of each device, and is, for example, the number of pages per block. The service writing method is set based on the specifications of each service. For example, there are a writing cycle and the number of data bytes written once.

  The area allocation process will be described. As shown in FIG. 3, the access control unit 12 allocates areas (block groups) for storing data of each service in the device (storage device 10) based on the used area size calculated for each service. . At this time, the device area is divided into a read / write area and a read area, and an area in the read / write area is assigned to a service that is mainly written, such as a vehicle diagnostic service, and reading is mainly performed like road information. In the service, an area is allocated in the Read area. The reason for dividing the area in this way is to prevent unnecessary writing / erasing from occurring in the data area mainly for reading due to leveling of the number of writing performed by the conventional wear leveling technique.

  This area allocation is performed in advance before starting the service. As with the used area size calculation process, for services that are determined to be installed at the design stage, area allocation processing is performed by a computer at the design stage, and areas for storing data for each service may be allocated. Good. In this case, the access control unit 12 performs the area allocation process only for a service that is added later in the distribution service or the like.

  The data writing process will be described. In each service, the access control unit 12 sequentially performs data writing / erasing for each unit in which the device can be most efficiently written in the area to which the service is allocated in accordance with the service writing method. In other words, rewriting data using the ring buffer method). For example, in the case of a NAND flash memory, the unit in which a device can be written most efficiently is a page unit.

  In this way, write / erase from the block next to the previously written block in the allocated area, and if the last block of the allocated area is reached, return to the first block in the area. By performing writing / erasing, the number of times of data rewriting of each block in the area becomes equal. Therefore, in order to write / erase data sequentially in units of allocated areas, wear leveling is performed in which the number of data rewrites in each block is managed and the number of write / erase times in each block is leveled. . In addition, since data is rewritten in the area allocated for each service unit, data rewriting to a block in an area allocated to another service is surely eliminated.

  Note that services that are mainly read, such as road information, are basically not written / erased except when data is updated. Therefore, data retention does not decrease for the Read area.

  An example in which a NAND flash memory is applied as the storage device 10 (device) will be described with reference to FIG. The service is a case of a vehicle diagnosis service. As a precondition, the upper limit number of writing is, for example, 100 times per page. The durability is 10 years. This takes into account the durable life (7 to 10 years) generally required for vehicles. As for the physical configuration of the device, one block is composed of 64 pages (one page is a data portion: 2048 bytes, a redundant portion: 64 bytes) (see FIG. 4). According to the specifications of the NAND flash memory, writing to the device is performed in units of one page. In the vehicle diagnosis service data writing method, 356 bytes per record is stored every 500 milliseconds in accordance with the specifications of the vehicle diagnosis service.

  Since one page of the NAND flash memory is 2048 bytes, one page can store five records of vehicle diagnosis service data (356 × 5 = 1780 bytes because one record is 356 bytes). Therefore, it is assumed that writing is performed in units of one page every 500 milliseconds × 5 records = 2.5 seconds. At this time, what is important is to write to the device by one page 2048 byte alignment. If you write to the device so that it spans two pages, ignoring the one page alignment, you need to read the first and second pages, align the data, and then write to the first and second pages. Writes occur (increase in the number of writes).

  The number of pages required for 10 years with the maximum number of write times of 100 per page is 10 years × 365 days × 24 hours × 60 minutes × 60 seconds ÷ 2.5 seconds ÷ 100 times = 1261440 pages. Here, the calculation is based on the assumption that the vehicle will be used for 365 days a day, 24 hours a day. The number of necessary blocks is 1261440 pages / 64 pages = 19710 blocks. The required use area size is 19710 blocks × (2048 bytes × 64 pages) = 2583429120 bytes (about 2.58 Gbytes). Furthermore, it is preferable to provide a margin of about several percent as the reliability (PPM) of the NAND flash memory to be used. For example, assuming a margin of 5%, 2.58 Gbyte × 1.05 = 2.709 Gbyte.

  When the use area size of the vehicle diagnosis service is calculated in advance as described above, an area corresponding to the size is assigned to the device in advance according to the use area size. Here, according to the required number of blocks calculated above, an area corresponding to the number of blocks may be allocated. Each time the vehicle travels, one page of vehicle diagnosis data (1780 bytes of data) is sequentially written into the allocated area every 2.5 seconds. Incidentally, for other services, similarly, the calculation of the used area size and the allocation of the area are performed in advance, and the data is sequentially written every time the vehicle travels.

  According to the storage device control apparatus 1, the use area size required for the durable years is obtained for each service, and wear leveling is performed in the allocated area for each use area size. The number of years can be secured. In particular, in the storage device control apparatus 1, a device is divided into a read / write area and a read area, a read / write area is allocated to a service in which writing is mainly performed (service with a high rewrite frequency), and a service in which reading is mainly performed (rewriting) Since the Read area is allocated to the service having a low frequency), it is possible to reliably prevent the data of the service that mainly writes data from being written / erased in the area allocated to the service that mainly reads data.

  Furthermore, according to the storage device control apparatus 1, the used area size is calculated for each service in consideration of the upper limit number of writes, and wear leveling is performed in the allocated area for each used area size. It is possible to prevent data from being written exceeding the upper limit number of times of writing. As a result, a decrease in data retention can be suppressed, and a write error can be suppressed. In particular, in the storage device control apparatus 1, data is sequentially written for each unit that can be most efficiently written to the device (for example, for each page that is the minimum block writing unit in the case of a NAND flash memory). The number of write / erase operations can be minimized.

  In addition, according to the storage device control apparatus 1, wear leveling can be realized by a simple method by sequentially writing data within the area allocated to each service. At this time, since the size that does not exceed the upper limit number of times of writing is obtained when the used area size is obtained, the number of times of writing / erasing for each block is counted, and the number of counts is determined by the number of times of threshold. No processing is necessary.

  As mentioned above, although embodiment which concerns on this invention was described, this invention is implemented in various forms, without being limited to the said embodiment.

  For example, in the present embodiment, the present invention is applied to a storage device control apparatus mounted on a vehicle, but it can be applied to various other systems using a storage device such as a flash memory or an SD memory card.

  In this embodiment, the present invention is applied to a flash memory and an SD memory, but may be applied to other storage devices such as an SSD.

  In this embodiment, the used area size is calculated based on the upper limit number of times of writing, the durable years, the physical configuration of the device, and the writing method of the service. To calculate.

  In this embodiment, wear leveling is performed by rewriting data sequentially for each area allocated to each service. However, wear leveling may be performed by other methods.

  DESCRIPTION OF SYMBOLS 1 ... Storage device control apparatus, 10 ... Memory | storage device, 11 ... IF part, 12 ... Access control part

Claims (3)

A hardware having a function of storing data, and performing wear leveling that equalizes the number of times of data rewriting for each of the plurality of blocks, and a plurality of blocks in which the period in which the stored data can be retained is shortened according to the number of times of data rewriting A storage device having leveling means,
Use area size calculation means for calculating a use area size necessary for a period in which data of the application must be held for each application that stores data in the storage device, and a block that divides the storage device into a plurality of block groups Having group dividing means;
The block group dividing means allocates a block group for the use area size calculated by the use area size calculation means for each application,
The wear leveling means performs wear leveling within a block group divided by the block group dividing means. The storage device according to claim 1 , wherein the storage device is a flash memory. The wear leveling means, for each block group, manages the data rewrite count of each block, according to claim 1 or claim 2, wherein selecting a block for performing data rewriting in accordance with the data rewrite count Storage device.

Images