CN102298973B - Radiation-resistant fault-protected storage device and radiation-resistant fault-protected method thereof - Google Patents
Radiation-resistant fault-protected storage device and radiation-resistant fault-protected method thereof Download PDFInfo
- Publication number
- CN102298973B CN102298973B CN201110147557.4A CN201110147557A CN102298973B CN 102298973 B CN102298973 B CN 102298973B CN 201110147557 A CN201110147557 A CN 201110147557A CN 102298973 B CN102298973 B CN 102298973B
- Authority
- CN
- China
- Prior art keywords
- code
- hamming
- decoding
- module
- ldpc
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 18
- 230000005855 radiation Effects 0.000 title description 14
- 230000003471 anti-radiation Effects 0.000 claims abstract description 14
- 238000001514 detection method Methods 0.000 claims description 48
- 239000011159 matrix material Substances 0.000 claims description 31
- 230000009897 systematic effect Effects 0.000 claims description 3
- 230000009466 transformation Effects 0.000 claims description 3
- 239000000523 sample Substances 0.000 claims 4
- 230000005055 memory storage Effects 0.000 claims 1
- 230000001052 transient effect Effects 0.000 abstract description 6
- 230000002401 inhibitory effect Effects 0.000 abstract 1
- 208000011580 syndromic disease Diseases 0.000 description 12
- 230000002787 reinforcement Effects 0.000 description 6
- 125000004122 cyclic group Chemical group 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 238000003491 array Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
Images
Landscapes
- Error Detection And Correction (AREA)
- Detection And Prevention Of Errors In Transmission (AREA)
- Detection And Correction Of Errors (AREA)
Abstract
Description
技术领域technical field
本发明涉及一种抗故障保护型存储装置及其保护方法。The invention relates to an anti-failure protection storage device and a protection method thereof.
背景技术Background technique
随着集成工艺尺寸的不断降低,集成电路对空间辐射环境和地面噪声环境越发地敏感,电路的正常工作状态受到严重地影响。错误修正码(ECC)是一种修正存储器中故障的常用方法。然而,随着存储器中相临单元之间的距离不断缩小,一次辐射事件造成多位翻转的几率大大地增加;同时由于ECC电路需要由编码器和译码器构成,这些组合电路会受到由辐射引起的单粒子瞬态效应的影响。因此,需要一种更为有效地的存储器故障保护(Fault-Secure)加固技术,既可以修正存储阵列中的多位错误,又可以修正ECC电路中的错误。With the continuous reduction of the size of the integrated process, the integrated circuit is more and more sensitive to the space radiation environment and the ground noise environment, and the normal working state of the circuit is seriously affected. Error Correcting Code (ECC) is a common method for correcting faults in memory. However, as the distance between adjacent cells in the memory continues to shrink, the probability of multiple bit flips caused by a radiation event greatly increases; at the same time, since the ECC circuit needs to be composed of an encoder and a decoder, these combined circuits will be affected by radiation. The influence of single-event transient effects induced. Therefore, there is a need for a more effective memory fault protection (Fault-Secure) hardening technology, which can not only correct multi-bit errors in the storage array, but also correct errors in the ECC circuit.
汉明码(Hamming codes)是ECC中最为常见的一种纠错技术,它可以修正一位、探测两位错误。汉明码具有编码结构简单,易于实现等特点,但是它无法修正多于两位的错误。欧氏几何低密度单奇偶校验码(EG-LDPC)是目前已知的唯一具有故障保护能力的ECC编码。EG-LDPC码是分组循环码的一种,它可以修正多个错误,同时具有高码增益和低误码率的特点,已经广泛应用于通信领域。EG-LDPC码还是一种大数逻辑可译码,相对于其它分组循环码的迭代译码方式,具有译码速度快的特点。具体译码方式如下:首先通过计算接收的数据向量和自身的奇偶校验矩阵的内积,得到α个奇偶校验总和;然后α个奇偶校验总和输入到大数逻辑门中进行判断;如果大数门的输出结果为1,则传输数据发生错误需要翻转相应的数据位;如果大数门的输出结果为0,则传输数据正确,保持不变。Hamming codes are the most common error correction technology in ECC, which can correct one bit and detect two bit errors. Hamming code has the characteristics of simple coding structure and easy implementation, but it cannot correct errors of more than two digits. Euclidean Geometry Low-Density Single Parity-Check Code (EG-LDPC) is the only ECC code known to be fail-safe. EG-LDPC code is a kind of block cyclic code, it can correct multiple errors, and has the characteristics of high code gain and low bit error rate, and has been widely used in the field of communication. EG-LDPC code is also a kind of large number logic decodable, compared with other iterative decoding methods of block cyclic codes, it has the characteristics of fast decoding speed. The specific decoding method is as follows: First, by calculating the inner product of the received data vector and its own parity check matrix, α parity check sums are obtained; then α parity check sums are input into the large number logic gate for judgment; if If the output result of the large number gate is 1, the corresponding data bit needs to be flipped if an error occurs in the transmitted data; if the output result of the large number gate is 0, the transmitted data is correct and remains unchanged.
一个ECC编码的故障保护能力是由其校正子(syndrome)Si构成的探测器电路决定的。校正子探测电路可以探测编码器、译码器和自身电路输出的信息位是否正确,如果译码正确则校正子为0,如果译码中出现错误则校正子不为0。对于一个字长为n的EG-LDPC码,它会产生n个校正子,用以产生探测信号并反馈回存储器,如果在编码器、译码器或探测器中出现无法修正的错误,则中断并重新发送出现错误的数据,从而使存储器具有了故障保护能力,修正方案如图1所示。The fault protection capability of an ECC code is determined by the detector circuit composed of its syndrome (syndrome) S i . The syndrome detection circuit can detect whether the information bits output by the encoder, decoder and its own circuit are correct. If the decoding is correct, the syndrome is 0, and if an error occurs in the decoding, the syndrome is not 0. For an EG-LDPC code with a word length of n, it will generate n syndromes, which are used to generate the detection signal and feed back to the memory. If an uncorrectable error occurs in the encoder, decoder or detector, it will be interrupted. And resend the wrong data, so that the memory has a fault protection capability, the correction scheme is shown in Figure 1.
但是,并不是所有的ECC编码都具有故障保护能力。设J为ECC码的最小距离(minimum distance),则ECC该的最大纠错能力E为(J-1)/2,最大探测能力D为J-1,错误图样的重量(weight)满足0<e≤J-1。在编码器、译码器和校正电路中发生的错误分别用Ee、Ed和Es标记,存储阵列中的错误为Em。对于普通ECC码需要满足Em≤E且Ee=Ed=0,而对于具有故障保护能力的ECC码则需满足Ee+Em≤E且Ee+Em+Ed+Es≤D。假定故障保护ECC中错误图样重量为Ee+Em+Ed=e,则校正子电路在可检测范围内发生错误的重量应满足Es≤J-1-e。要保证校正子电路中每个错误只影响一个校正子位,实现对校正子电路自身出现错误的探测,此时校正子的重量至少需要为J-e。目前已经证明EG-LDPC码的校正子重量可以满足大于J-e,因此具备故障保护的能力。However, not all ECC encodings are failsafe. Let J be the minimum distance of the ECC code, then the maximum error correction capability E of the ECC code is (J-1)/2, the maximum detection capability D is J-1, and the weight of the error pattern satisfies 0< e≤J-1. The errors occurring in the encoder, decoder and correction circuit are marked with E e , E d and E s respectively, and the errors in the storage array are E m . For ordinary ECC codes, E m ≤ E and E e = E d = 0 must be satisfied, while for ECC codes with fault protection capabilities, E e + E m ≤ E and E e + E m + E d + E s must be satisfied ≤D. Assuming that the error pattern weight in the fail-safe ECC is E e +E m +E d =e, the error weight of the syndrome circuit within the detectable range should satisfy E s ≤ J-1-e. To ensure that each error in the syndrome circuit affects only one syndrome bit, and realize the detection of errors in the syndrome circuit itself, the weight of the syndrome needs to be at least Je. It has been proven that the syndrome weight of the EG-LDPC code can be greater than Je, so it has the ability of fault protection.
由于EG-LDPC码需要较多的冗余位存放编码信息,而存储单元又占据了整个存储器的绝大部分面积,因此EG-LDPC码的故障保护方案会带来庞大的面积开销,增加了芯片的成本。此外,某些码段的EG-LDPC需要多步大数逻辑实现译码,从而带来过大的延迟开销,并不适合一些高速存储器的应用。Since EG-LDPC codes require more redundant bits to store coding information, and the storage unit occupies most of the area of the entire memory, the fault protection scheme of EG-LDPC codes will bring huge area overhead and increase the chip cost. the cost of. In addition, the EG-LDPC of some code segments requires multi-step large number logic to realize decoding, which brings excessive delay overhead and is not suitable for some high-speed memory applications.
发明内容Contents of the invention
本发明未解决现有存储阵列中出现的多位翻转又可以抑制编码器、译码器等组合电路中出现的单粒子瞬态效应,及现有EG-LDPC码因需要较多的冗余位存放编码信息而带来庞大的面积开销,增加芯片成本的问题;而提出了一种抗辐射故障保护型存储装置及其抗辐射故障保护方法。The present invention does not solve the multi-bit inversion that occurs in the existing storage array, but also can suppress the single-event transient effect that occurs in combination circuits such as encoders and decoders, and the existing EG-LDPC code requires more redundant bits. The storage of encoded information brings huge area overhead and increases the cost of the chip; and a radiation-resistant fault-protection type storage device and a radiation-resistant fault protection method thereof are proposed.
抗辐射故障保护型存储装置,它包括混合码编码组件和混合码译码组件;所述混合码编码组件由EG-LDPC码编码模块和汉明码编码模块组成;EG-LDPC码编码模块的信息编码数据输入端和汉明码编码模块的信息编码数据输入端同时与外部信息数据输出端相连;EG-LDPC码编码模块的EG-LDPC码编码数据输出端与存储阵列模块的EG-LDPC码编码数据输入端相连;汉明码编码模块的汉明码编码数据输出端与存储阵列模块的汉明码编码数据输入端相连;所述混合码译码组件由EG-LDPC码译码模块和汉明码译码模块组成;EG-LDPC码译码模块的EG-LDPC码译码数据输入端与存储阵列模块的EG-LDPC码译码数据输出端相连;汉明码译码模块的汉明码译码数据输入端与存储阵列模块的汉明码译码数据输出端相连;它还包括编码错误码探测组件和译码错误码探测组件;所述编码错误码探测组件由EG-LDPC码编码错误码探测模块和汉明码编码错误码探测模块组成;Anti-radiation fault protection type storage device, it comprises mixed code coding component and mixed code decoding component; Described mixed code coding component is made up of EG-LDPC code coding module and Hamming code coding module; Information coding of EG-LDPC code coding module The data input end and the information encoding data input end of the Hamming code encoding module are connected to the external information data output end at the same time; the EG-LDPC code encoding data output end of the EG-LDPC code encoding module is connected to the EG-LDPC code encoding data input of the storage array module The terminal is connected; the Hamming code encoding data output end of the Hamming code encoding module is connected with the Hamming code encoding data input end of the storage array module; the mixed code decoding assembly is composed of an EG-LDPC code decoding module and a Hamming code decoding module; The EG-LDPC code decoding data input terminal of the EG-LDPC code decoding module is connected to the EG-LDPC code decoding data output terminal of the storage array module; the Hamming code decoding data input terminal of the Hamming code decoding module is connected to the storage array module The Hamming code decoding data output end is connected; It also includes coding error code detection component and decoding error code detection component; Described coding error code detection component is detected by EG-LDPC code coding error code detection module and Hamming code coding error code detection Module composition;
所述EG-LDPC码编码错误码探测模块的EG-LDPC码编码错误码探测数据输入端与EG-LDPC码编码模块的EG-LDPC码编码数据输出端相连;所述汉明码编码错误码探测模块的汉明码编码错误码探测数据输入端与汉明码编码模块的汉明码编码数据输出端相连;The EG-LDPC code encoding error code detection data input end of the EG-LDPC code encoding error code detection module is connected with the EG-LDPC code encoding data output end of the EG-LDPC code encoding module; the Hamming code encoding error code detection module The Hamming code encoding error code detection data input end of the Hamming code encoding module is connected to the Hamming code encoding data output end;
所述译码错误码探测组件由EG-LDPC码译码错误码探测模块和汉明码译码错误码探测模块组成;Described decoding error code detection assembly is made up of EG-LDPC code decoding error code detection module and Hamming code decoding error code detection module;
所述EG-LDPC码译码错误码探测模块的EG-LDPC码译码错误码探测数据输入端与EG-LDPC码译码模块的EG-LDPC码译码数据输出端相连;所述汉明码译码错误码探测模块的汉明码译码错误码探测数据输入端与汉明码译码模块的汉明码译码数据输出端相连;The EG-LDPC code decoding error code detection data input end of the EG-LDPC code decoding error code detection module is connected with the EG-LDPC code decoding data output end of the EG-LDPC code decoding module; the Hamming code decoding The Hamming code decoding error code detection data input end of the code error code detection module is connected with the Hamming code decoding data output end of the Hamming code decoding module;
所述EG-LDPC码译码错误码探测模块的EG-LDPC码译码伴随数据输入端同时与EG-LDPC码译码模块的EG-LDPC码译码伴随数据输出端、汉明码译码模块的汉明码译码伴随数据输出端和汉明码译码错误码探测模块的汉明码译码伴随数据输入端与相连。The EG-LDPC code decoding accompanying data input end of the EG-LDPC code decoding error code detection module is simultaneously with the EG-LDPC code decoding accompanying data output end of the EG-LDPC code decoding module, and the Hamming code decoding module The Hamming code decoding accompanying data output terminal is connected with the Hamming code decoding accompanying data input terminal of the Hamming code decoding error code detection module.
采用上述的抗辐射故障保护型存储装置的抗辐射故障保护方法,它由以下几个步骤组成:The anti-radiation fault protection method of the above-mentioned anti-radiation fault protection storage device is composed of the following steps:
步骤1:根据需要加固存储器的数据宽度N,选择码字为(n1,k1)的EG-LDPC码和码字为(n2,k2)的汉明码;其中n1和k1分别为EG-LDPC码的码长和数据宽度,n2和k2分别为汉明码的码长和数据宽度;Step 1: According to the data width N of the hardened memory, select the EG-LDPC code with the code word (n1, k1) and the Hamming code with the code word (n2, k2); where n1 and k1 are the EG-LDPC code Code length and data width, n2 and k2 are the code length and data width of Hamming code respectively;
步骤2:将步骤1选择的码长为n1的EG-LDPC码分割成M个部分,M的取值与汉明码的码长n2相等,所述M个部分中的每一部分的长度至少2位;Step 2: Divide the EG-LDPC code with code length n1 selected in
步骤3:把码长为n2的汉明码均匀地插入到EG-LDPC码分割的M个部分的间隔中,使汉明码的每一个字节在物理布局上都是分隔的;若汉明码和EG-LDPC码的码长和信息位满足等式1和2,即构成一个数据宽度为N,码长为n1+n2的混合码;Step 3: Insert the Hamming code with a code length of n2 evenly into the interval of the M parts divided by the EG-LDPC code, so that each byte of the Hamming code is physically separated; if the Hamming code and EG -The code length and information bits of the LDPC code satisfy
k1+k2=N 等式2k1+k2=
步骤4:通过故障保护约束算法来确保混合码的故障保护特性;设J为错误修正码的最小距离,e为错误图样的重量,对系统形式的校验矩阵Step 4: Ensure the fail-safe characteristics of the hybrid code through the fail-safe constraint algorithm; let J be the minimum distance of the error-correcting code, e be the weight of the error pattern, and check matrix of the system form
进行行初等变换,得到具有故障保护能力的校验矩阵HFS;Carry out row elementary transformation to obtain the parity check matrix H FS with fault protection capability;
所述具有故障保护能力的校验矩阵HFS满足下列条件:The check matrix HFS with fault protection capability satisfies the following conditions:
a、每一列中至少包含J个1;a. Each column contains at least J 1s;
b、故障保护校验矩阵HFS中1的总个数Z保持在"(J+1)×矩阵列数≥Z≥J×矩阵列数"之间;b. The total number Z of 1s in the failsafe parity check matrix H FS is kept between "(J+1)×number of matrix columns≥Z≥J×matrix column number';
c、故障保护校验矩阵HFS中每一行中1的个数都相等;c. The number of 1s in each row in the failsafe parity check matrix HFS is equal;
步骤5:利用EG-LDPC码校验矩阵和步骤4得到的汉明码校验矩阵HFS,通过与接收和发送的数据位执行相应的向量矩阵乘法获得混合码的编码器的参数、译码器的参数和探测器的参数;所述混合码的编码器、译码器和探测器连接构成抗辐射故障保护型存储装置。Step 5: Use the EG-LDPC code check matrix and the Hamming code check matrix H FS obtained in
本发明所述的抗辐射故障保护型存储装置具有低面积和延迟开销的特性,适用于同时抑制存储阵列中的多位翻转与编码器、译码器等组合电路中的单粒子瞬态效应。The anti-radiation fault protection storage device of the invention has the characteristics of low area and delay overhead, and is suitable for simultaneously suppressing multi-bit inversion in a storage array and single event transient effects in combined circuits such as encoders and decoders.
附图说明Description of drawings
图1为ECC编码的故障保护装置结构示意图;图2为本发明的模块结构示意图;图3为具体实施方式二所述方法中混合码字的结构,其中M代表汉明码,E代表EG-LDPC码,A代表错误间隔,B代表每一部分的长度。Fig. 1 is the fault protection device structural representation of ECC coding; Fig. 2 is the module structural representation of the present invention; Fig. 3 is the structure of mixed code word in the method described in specific embodiment two, wherein M represents Hamming code, and E represents EG-LDPC code, A represents the error interval, and B represents the length of each part.
具体实施方式Detailed ways
具体实施方式一:结合图2说明本实施方式,本实施方式所述的抗辐射故障保护型存储装置,它包括混合码编码组件2和混合码译码组件3;所述混合码编码组件2由EG-LDPC码编码模块2-1和汉明码编码模块2-2组成;EG-LDPC码编码模块2-1的信息编码数据输入端和汉明码编码模块2-2的信息编码数据输入端同时与外部信息数据输出端相连;EG-LDPC码编码模块2-1的EG-LDPC码编码数据输出端与存储阵列模块1的EG-LDPC码编码数据输入端相连;汉明码编码模块2-2的汉明码编码数据输出端与存储阵列模块1的汉明码编码数据输入端相连;所述混合码译码组件3由EG-LDPC码译码模块3-1和汉明码译码模块3-2组成;EG-LDPC码译码模块3-1的EG-LDPC码译码数据输入端与存储阵列模块1的EG-LDPC码译码数据输出端相连;汉明码译码模块3-2的汉明码译码数据输入端与存储阵列模块1的汉明码译码数据输出端相连;它还包括编码错误码探测组件4和译码错误码探测组件5;所述编码错误码探测组件4由EG-LDPC码编码错误码探测模块4-1和汉明码编码错误码探测模块4-2组成;Specific Embodiment 1: This embodiment is described in conjunction with FIG. 2 . The anti-radiation fault protection storage device described in this embodiment includes a mixed
所述EG-LDPC码编码错误码探测模块4-1的EG-LDPC码编码错误码探测数据输入端与EG-LDPC码编码模块2-1的EG-LDPC码编码数据输出端相连;所述汉明码编码错误码探测模块4-2的汉明码编码错误码探测数据输入端与汉明码编码模块2-2的汉明码编码数据输出端相连;The EG-LDPC code encoding error code detection data input end of the EG-LDPC code encoding error code detection module 4-1 is connected with the EG-LDPC code encoding data output end of the EG-LDPC code encoding module 2-1; The Hamming code coding error code detection data input end of the plain code coding error code detection module 4-2 is connected with the Hamming code coding data output end of the Hamming code coding module 2-2;
所述译码错误码探测组件5由EG-LDPC码译码错误码探测模块5-1和汉明码译码错误码探测模块5-2组成;Described decoding error
所述EG-LDPC码译码错误码探测模块5-1的EG-LDPC码译码错误码探测数据输入端与EG-LDPC码译码模块3-1的EG-LDPC码译码数据输出端相连;所述汉明码译码错误码探测模块5-2的汉明码译码错误码探测数据输入端与汉明码译码模块3-2的汉明码译码数据输出端相连;The EG-LDPC code decoding error code detection data input end of the EG-LDPC code decoding error code detection module 5-1 is connected with the EG-LDPC code decoding data output end of the EG-LDPC code decoding module 3-1 The Hamming code decoding error code detection data input end of the Hamming code decoding error code detection module 5-2 is connected to the Hamming code decoding data output end of the Hamming code decoding module 3-2;
所述EG-LDPC码译码错误码探测模块5-1的EG-LDPC码译码伴随数据输入端同时与EG-LDPC码译码模块3-1的EG-LDPC码译码伴随数据输出端、汉明码译码模块3-2的汉明码译码伴随数据输出端和汉明码译码错误码探测模块5-2的汉明码译码伴随数据输入端与相连。所述EG-LDPC码即为欧氏几何低密度单奇偶校验码。The EG-LDPC code decoding accompanying data input terminal of the EG-LDPC code decoding error code detection module 5-1 is simultaneously connected with the EG-LDPC code decoding accompanying data output terminal of the EG-LDPC code decoding module 3-1, The Hamming code decoding accompanying data output terminal of the Hamming code decoding module 3-2 is connected with the Hamming code decoding accompanying data input terminal of the Hamming code decoding error code detection module 5-2. The EG-LDPC code is the Euclidean geometry low density single parity check code.
抗辐射故障保护型存储装置工作原理如下:数据通过混合码编码组件2存入存储阵列模块1中,如果辐射事件影响存储阵列模块1中多个数据位,使数据的状态发生改变,那么数据在读出存储阵列模块1的过程中,通过混合码译码组件3可以修正这些数据。如果错误出现在混合码编码组件2、混合码译码组件3、编码错误码探测组件4和译码错误码探测组件5中,那么可以通过编码错误码探测组件4和译码错误码探测组件5检测并随之修正错误。本发明所述的抗辐射故障保护型存储装置具有低面积和延迟开销的特性,适用于同时抑制存储阵列中的多位翻转与编码器、译码器等组合电路中的单粒子瞬态效应。The working principle of the anti-radiation fault protection type storage device is as follows: data is stored in the
具体实施方式二:采用具体实施方式一所述的抗辐射故障保护型存储装置的抗辐射故障保护方法,它由以下几个步骤组成:Embodiment 2: The radiation-resistant fault protection method of the radiation-resistant fault-protected storage device described in
步骤1:根据需要加固存储器的数据宽度N,选择码字为(n1,k1)的EG-LDPC码和码字为(n2,k2)的汉明码;其中n1和k1分别为EG-LDPC码的码长和数据宽度,n2和k2分别为汉明码的码长和数据宽度;Step 1: According to the data width N of the reinforced memory, select the EG-LDPC code with the code word (n1, k1) and the Hamming code with the code word (n2, k2); where n1 and k1 are the EG-LDPC code Code length and data width, n2 and k2 are the code length and data width of Hamming code respectively;
步骤2:将步骤1选择的码长为n1的EG-LDPC码分割成M个部分,M的取值与汉明码的码长n2相等,所述M个部分中的每一部分的长度至少2位;Step 2: Divide the EG-LDPC code with code length n1 selected in
步骤3:把码长为n2的汉明码均匀地插入到EG-LDPC码分割的M个部分的间隔中,使汉明码的每一个字节在物理布局上都是分隔的;若汉明码和EG-LDPC码的码长和信息位满足等式1和2,即构成一个数据宽度为N,码长为n1+n2的混合码;Step 3: Insert the Hamming code with a code length of n2 evenly into the interval of the M parts divided by the EG-LDPC code, so that each byte of the Hamming code is physically separated; if the Hamming code and EG -The code length and information bits of the LDPC code satisfy
k1+k2=N 等式2k1+k2=
步骤4:通过故障保护约束算法来确保混合码的故障保护特性;设J为错误修正码的最小距离,e为错误图样的重量,对系统形式的校验矩阵Step 4: Ensure the fail-safe characteristics of the hybrid code through the fail-safe constraint algorithm; let J be the minimum distance of the error-correcting code, e be the weight of the error pattern, and check matrix of the system form
进行行初等变换,得到具有故障保护能力的校验矩阵Hfs;Carry out row elementary transformation to obtain the parity check matrix H fs with fault protection capability;
所述具有故障保护能力的校验矩阵Hfs满足下列条件:The check matrix H fs with fault protection capability satisfies the following conditions:
a、每一列中至少包含J个1;a. Each column contains at least J 1s;
b、故障保护校验矩阵Hfs中1的总个数Z保持在"(J+1)×矩阵列数≥Z≥J×矩阵列数"之间;b. The total number Z of 1s in the failsafe parity check matrix Hfs is kept between "(J+1)×number of matrix columns≥Z≥J×matrix column number';
c、故障保护校验矩阵Hfs中每一行中1的个数都相等;c. The number of 1s in each row in the failsafe check matrix H fs is equal;
步骤6:利用EG-LDPC码校验矩阵和步骤5得到的汉明码校验矩阵Hfs,通过与接收和发送的数据位执行相应的向量矩阵乘法获得混合码的编码器、译码器和探测器;所述混合码的编码器、译码器和探测器连接构成抗辐射故障保护型存储装置。Step 6: Use the EG-LDPC code check matrix and the Hamming code check matrix H fs obtained in
目的:混合码编码器对写入数据进行编码,编码后的数据存入存储器中。混合码译码器对读出存储器的数据进行译码,如果辐射事件影响存储器中多个数据位,使数据的状态发生改变,那么通过混合码译码器可以修正这些多位翻转。混合码探测器可以探测并随之修正编码器、译码器和探测器中出现的单粒子翻转。Purpose: The mixed code encoder encodes the written data, and the encoded data is stored in the memory. The mixed code decoder decodes the data read out of the memory. If the radiation event affects multiple data bits in the memory and changes the state of the data, then these multi-bit inversions can be corrected by the mixed code decoder. Hybrid code detectors can detect and subsequently correct single-event upsets in encoders, decoders, and detectors.
通过上述方法得到的混合码所构成的故障保护储器系统具有低面积和延迟开销的特性,适用于同时抑制存储阵列中的多位翻转与编码器、译码器等组合电路中的单粒子瞬态效应。The fault-protected memory system composed of the mixed code obtained by the above method has the characteristics of low area and delay overhead, and is suitable for simultaneously suppressing multi-bit flips in memory arrays and single event transients in combinational circuits such as encoders and decoders. state effect.
具体实施方式三:本实施方式与具体实施方式二不同点在于把一码字定义为(n,k,t)的形式,其中n是码长,k是信息位,t代表纠错能力。EG-LDPC码可以对存储阵列模块1中通常使用的数据位提供较强的纠错能力,但是需要通过双倍的冗余和多步译码实现,会对整个存储阵列模块1带来过大的面积和延迟开销。此外,过高的错误修正能力(例如15位)对通常的辐射环境来说是不需要的。混合码的加固方案Embodiment 3: The difference between this embodiment and
混合码的加固方案和EG-LDPC码的加固方案对比如表I所示。表中第一行为EG-LDPC码的加固方案,第二、三行分别为选取的低权重或译码级数较少的EG-LDPC码和相应的汉明码,第三行为混合码加固方案。从表I中可以看到,混合码的加固方案冗余位和译码步骤均有明显地降低。其它组成和连接方式与具体实施方式二相同。The comparison between the reinforcement scheme of the hybrid code and the reinforcement scheme of the EG-LDPC code is shown in Table I. The first line in the table is the reinforcement scheme of the EG-LDPC code, the second and third lines are the selected EG-LDPC codes with low weight or fewer decoding stages and the corresponding Hamming codes respectively, and the third line is the mixed code reinforcement scheme. It can be seen from Table I that the redundant bits and decoding steps of the reinforcement scheme of the hybrid code are significantly reduced. Other compositions and connection methods are the same as those in the second embodiment.
以上内容是结合具体的优选实施方式对本发明所作的进一步详细说明,不能认定本发明的具体实施只局限于这些说明。对于本所属技术领域的普通技术人员来说,在不脱离本发明构思的前提下,还可以做出若干简单推演或替换,都应当视为属于本发明所提交的权利要求书确定的专利保护范围。The above content is a further detailed description of the present invention in conjunction with specific preferred embodiments, and it cannot be assumed that the specific implementation of the present invention is limited to these descriptions. For those of ordinary skill in this technical field, without departing from the concept of the present invention, some simple deduction or replacement can also be made, which should be regarded as belonging to the scope of patent protection determined by the claims submitted by the present invention .
Claims (2)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110147557.4A CN102298973B (en) | 2011-06-02 | 2011-06-02 | Radiation-resistant fault-protected storage device and radiation-resistant fault-protected method thereof |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110147557.4A CN102298973B (en) | 2011-06-02 | 2011-06-02 | Radiation-resistant fault-protected storage device and radiation-resistant fault-protected method thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102298973A CN102298973A (en) | 2011-12-28 |
| CN102298973B true CN102298973B (en) | 2014-02-26 |
Family
ID=45359321
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110147557.4A Expired - Fee Related CN102298973B (en) | 2011-06-02 | 2011-06-02 | Radiation-resistant fault-protected storage device and radiation-resistant fault-protected method thereof |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102298973B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103971747B (en) * | 2013-01-24 | 2018-05-01 | 深圳市国微电子有限公司 | A kind of programmable memory of anti-fuse type |
| CN103165193B (en) * | 2013-03-01 | 2016-03-02 | 中国空间技术研究院 | Low redundancy reinforcement means and the circuit arrangement of storer two bit-errors are revised in detection |
| CN103812472B (en) * | 2014-03-03 | 2016-06-01 | 哈尔滨工业大学 | The triggering device of anti-single particle transient state effect |
| CN109903806A (en) * | 2019-04-01 | 2019-06-18 | 江苏华存电子科技有限公司 | A kind of double mode error detection memory and double mode error-detecting method |
| CN110111826B (en) * | 2019-04-16 | 2021-04-13 | 北京空间飞行器总体设计部 | A spacecraft using SRAM-type FPGA dual-port RAM anti-single-event overturning reinforcement device |
| CN110209524B (en) * | 2019-06-18 | 2022-12-06 | 哈尔滨工业大学 | ECC decoder reinforcing method for resisting single-event transient effect |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101814922A (en) * | 2009-02-23 | 2010-08-25 | 国际商业机器公司 | Multi-bit error correcting method and device based on BCH (Broadcast Channel) code and memory system |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040163030A1 (en) * | 2003-02-13 | 2004-08-19 | International Business Machines Corporation | Iterative error correcting system |
-
2011
- 2011-06-02 CN CN201110147557.4A patent/CN102298973B/en not_active Expired - Fee Related
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101814922A (en) * | 2009-02-23 | 2010-08-25 | 国际商业机器公司 | Multi-bit error correcting method and device based on BCH (Broadcast Channel) code and memory system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102298973A (en) | 2011-12-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11740960B2 (en) | Detection and correction of data bit errors using error correction codes | |
| Neale et al. | A new SEC-DED error correction code subclass for adjacent MBU tolerance in embedded memory | |
| Neale et al. | Adjacent-MBU-tolerant SEC-DED-TAEC-yAED codes for embedded SRAMs | |
| CN102915768B (en) | The fault-tolerant devices and methods therefor of triplication redundancy based on EDAC module to memory | |
| CN102298973B (en) | Radiation-resistant fault-protected storage device and radiation-resistant fault-protected method thereof | |
| US8806295B2 (en) | Mis-correction and no-correction rates for error control | |
| Tambatkar et al. | Error detection and correction in semiconductor memories using 3D parity check code with hamming code | |
| US10291258B2 (en) | Error correcting code for correcting single symbol errors and detecting double bit errors | |
| Sridevi et al. | Implementation of error correction techniques in memory applications | |
| Li et al. | Efficient implementations of 4-bit burst error correction for memories | |
| US8875002B1 (en) | Low cost adjacent double error correcting code | |
| Liu et al. | Low redundancy matrix-based codes for adjacent error correction with parity sharing | |
| US10567007B2 (en) | Device and method of processing a data word using checkbits | |
| Saiz-Adalid et al. | Modified Hamming codes to enhance short burst error detection in semiconductor memories (short paper) | |
| Sim et al. | Design of two interleaved error detection and corrections using Hsiao code and CRC | |
| Badack et al. | Modified DEC BCH codes for parallel correction of 3-bit errors comprising a pair of adjacent errors | |
| Rao et al. | Encoder and adaptive decoder for a (15, 6, 2) DEC-TED BCH code | |
| Maity et al. | FPGA-based low delay adjacent triple-bit error correcting codec | |
| Kim et al. | Hamming product code with iterative process for NAND flash memory controller | |
| Hamburg et al. | Unraveling codes: fast, robust, beyond-bound error correction for DRAM | |
| Sanilkumar et al. | Efficient Error Correcting Mechanism for Memories used in Radiated Environment | |
| Xiao et al. | Hardened design based on advanced orthogonal Latin code against two adjacent multiple bit upsets (MBUs) in memories | |
| Jayarani et al. | A novel fault detection and correction technique for memory applications | |
| Pushpalatha et al. | Error Detection and Correction using Decimal Matrix Algorithm | |
| Ahmed et al. | Double Adjacent Error Correction Codes for Ultra-Fast Cache Memories |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140226 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |