KR101077972B1 - Apparatus and method for parallel-processing of AES-CCM using general purpose multi-core processor - Google Patents

Abstract

The present invention relates to an AES-CCM parallel processing technology. The AES-CCM parallel processing apparatus according to the present invention includes: a core allocator configured to allocate a core of a general-purpose multicore processor to a task to be processed by the AES-CCM; And a data parallelization unit configured to perform data parallelism with respect to the processing target task in response to the number of the assigned cores, wherein the general-purpose multicore processor is configured to assign the data parallelized processing target task to the allocated parallel processing unit. Performing in parallel through the cores ensures data confidentiality and integrity simultaneously, while improving processing speed and resource efficiency and providing processor compatibility.

Description

Apparatus and method for parallel-processing of AES-CCM using general purpose multi-core processor}

The present invention relates to an AES-CCM parallel processing technology, and more particularly, by parallel processing the AES-CCM using a general-purpose multicore processor, thereby ensuring data confidentiality and integrity simultaneously, as well as processing speed and resource efficiency. AES-CCM parallel processing apparatus and method for improving and having processor compatibility.

As the digital information society is advanced and electronic commerce is activated, cryptography and authentication technologies are now recognized as key technologies for the safety and reliability of social and economic activities based on the Internet, and the protection of user privacy. In addition, with the recent increase in the use of large amounts of data such as MPEG video streams, a problem of protecting large amounts of data information using such encryption / authentication techniques has been raised.

In order to protect large data information, it is necessary to additionally apply a message authentication method such as a keyed-hash message authentication code (HMAC) as well as a method of ensuring data confidentiality. In other words, the encryption algorithm that can authenticate the message should be applied to ensure the integrity of large data at the same time.

As such, there is a method using a block encryption operation mode as a method for simultaneously ensuring the confidentiality and integrity of data.

Adopted by the National Institute of Standards and Technology (NIST) as the next generation of international standard cryptography, the Advanced Encryption Standard (AES) defines five operating modes as standards to ensure data integrity and confidentiality. That is, ECB (Electronic Code Book), CBC (Cipher Block Chaining), CFB (Cipher FeedBack), OFD (Output FeedBack) and CTR (Counter) are prescribed.

AES-CCM (Counter with CBC-MAC), which combines CTR mode and CBC-MAC mode among the various standard operating modes, combines the functions of the two combined modes to provide encryption and message integrity in one solution. . That is, the confidentiality function, the key management function, and the message integrity function can be simultaneously performed. The AES-CCM is commonly adopted by IEEE 802.11i, which is a wireless LAN security standard, and WPA, and is optional in WPA and mandatory in WPA2.

On the other hand, the AES-CCM has the advantage of showing a high reliability because it uses the two standard techniques already verified as mentioned above. Also, in general, unlike using a different key for authentication and encryption, there is an advantage that only one key can be used. In addition, AES-CCM can be encrypted without additional ciphertext overhead even when data that requires only integrity verification without encryption has the advantage of easy data handling. Furthermore, AES-CCM has a merit that it can be implemented with a relatively small code because it performs both encryption and decryption using only one encryption function. These advantages enable flexible application of AES-CCM over other modes of operation.

However, the existing technology has a problem in that the encryption process and computing time are delayed and the processing speed is slow in using a single core in processing AES-CCM which requires two operations for encryption and message integrity.

In addition, the existing technology has a problem in that it uses a dedicated hardware chip in consideration of the AES-CCM processing speed, so that resource efficiency is lowered and compatibility is impaired.

Accordingly, the first technical problem to be solved by the present invention is to parallelize the AES-CCM using a general-purpose multicore processor, thereby ensuring data confidentiality and integrity simultaneously, as well as improving processing speed and resource efficiency and processor compatibility. To provide an AES-CCM parallel processing unit having a.

The second technical problem to be solved by the present invention is to parallelize the AES-CCM using a general-purpose multi-core processor, which ensures data confidentiality and integrity simultaneously, improves processing speed and resource efficiency, and has processor compatibility. To provide AES-CCM parallel processing method.

In order to solve the first technical problem as described above, the present invention, the core allocation unit for allocating the core of the general-purpose multi-core processor for the task (task) of the AES-CCM; And a data parallelization unit configured to perform data parallelism with respect to the processing target task in response to the number of the assigned cores, wherein the general-purpose multicore processor is configured to assign the data parallelized processing target task to the allocated parallel processing unit. An AES-CCM parallel processing apparatus using a general-purpose multicore process that executes in parallel through a core is provided.

The core allocator may allocate two or more of the cores of the general-purpose multicore process to the processing target task when there is no dependency between the data of the processing target task.

The core allocator allocates one core among cores of the general purpose multicore process to the processing target task when there is a dependency between data of the processing target task.

The core allocator may allocate two or more cores when the task to be processed is a counter (CTR) among the cores of the general-purpose multicore process, and the task to be processed is a CBC-MAC (Cipher Block). In case of Chaining-Message Authentication Code), one core is allocated.

The core allocator allocates an entire core of the general purpose multicore process when the task to be processed is the CTR.

In one embodiment, the general-purpose multicore processor, when performing the processing for the CTR and the CBC-MAC, performs the processing for the CTR or the CBC-MAC first.

In one embodiment, the data parallelization unit does not perform the data parallelization when there is a dependency between data of the processing target task.

The data parallelization unit does not perform the data parallelization when the task to be processed is the CBC-MAC.

The AES-CCM parallel processing apparatus may further include a task parallelization unit configured to perform task parallelism on CTR and CBC-MAC among tasks to be processed by the AES-CCM.

In one embodiment, the core allocator allocates one core for the CBC-MAC and one or more cores for the CTR among the cores of the general-purpose multicore process.

In one embodiment, the data parallelization unit performs the data parallelization only for the CTR.

In one embodiment, the general purpose multicore processor performs the processing for the CTR and the CBC-MAC in parallel.

In order to solve the second technical problem as described above, the present invention, in a computer system using a general-purpose multi-core processor, the core allocation for allocating the core of the general-purpose multi-core processor to the processing target task (AES-CCM) step; In the computer system, performing data parallelism on the task to be processed in correspondence to the number of allocated cores; And a processing step of processing the data-parallelized task to be processed in parallel through the allocated cores in the general-purpose multicore processor.

In another aspect of the present invention, the AES-CCM parallel processing method according to the present invention is a task for CTR and CBC-MAC among the tasks to be processed in the AES-CCM in a computer system using a general-purpose multicore processor. Task parallelism performing task parallelism; In the computer system, a core allocation step of allocating one core for the CBC-MAC and one or more cores for the CTR among the cores of the general-purpose multicore process; In the computer system, performing data parallelism on the CTR in correspondence with the allocated number of cores; And in the general purpose multicore processor, processing the data parallelized CTR and the CBC-MAC in parallel through the assigned core.

In one embodiment, the invention may be embodied in a computer readable recording medium having recorded thereon a program for executing the method on a computer.

In one embodiment, the program is programmed through POSIX Threads (Pthreads), which is a standard application programming interface (API) used for writing software that runs in parallel.

The present invention provides the advantage of improving processing speed and resource efficiency as well as ensuring data confidentiality and integrity simultaneously by parallelizing AES-CCM.

In addition, using a general-purpose multicore processor provides the advantage of enabling cost-effective solution development without the need for separate hardware.

Furthermore, using Pthreads provides the advantage of being compatible with a variety of CPUs.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings in order to clarify a solution of the technical problem of the present invention. However, in the following description of the present invention, if it is determined that the description of the related known technology may obscure the gist of the present invention, the description thereof will be omitted. In addition, terms to be described later are terms defined in consideration of functions in the present invention, which may vary according to intentions or customs of users, operators, and the like. Therefore, the definition should be based on the contents throughout this specification.

Figure 1 shows a block diagram of an AES-CCM parallel processing apparatus using a general-purpose multicore process according to the present invention.

As shown in FIG. 1, the AES-CCM parallel processing apparatus 100 according to the present invention includes a core allocation unit 120, a data parallelization unit 130, and a general purpose multicore processor 140, and task parallelization. The unit 110 may further include.

First, the AES-CCM processing process performed by the AES-CCM parallel processing apparatus 100 is analyzed.

2 is a flowchart illustrating the processing of the AES-CCM.

As shown in FIG. 2, the processing of the AES-CCM includes a process of performing message encryption / authentication using a key initialization process (S210), a message header verification process (S220), a CTR, and a CBC-MAC (S230). ), And verifying the MAC value through MAC calculation (S240).

In the key initialization process (S210) and the header verification process (S220) of the message, since small data is processed, it does not significantly affect the overall execution time. In addition, since the MAC calculation process (S240) is a process of performing only XOR operation on the encrypted block, there is no reason for the parallel processing. In practical implementations, when parallelizing tasks with low data throughput, complex processing due to parallelism may impair efficiency.

On the other hand, the message encryption / authentication process (S230), the calculation to encrypt the message using the CTR (Counter) and the calculation to generate the MAC block using the Cipher Block Chaining-Message Authentication Code Protocol (CBC-MAC) do. The message encryption / authentication process (S230) occupies most of the entire execution time of the AES-CCM since the entire message must be processed as many as the number of encryption / authentication blocks. In addition, since both processes use AES encryption and process messages of the same size, it takes almost similar execution time.

Accordingly, the present invention is to efficiently parallelize the message encryption / authentication process (S230), particularly in AES-CCM.

3 shows a process of the CTR operation mode.

As shown in FIG. 3, the CTR increases the block made using a random number and a counter to a constant value, and thus uses the block cipher as a stream cipher. In particular, the key and nonce can be precomputed, allowing data to be encrypted in real time, with no dependencies between the data in the process. As will be described again below, the apparatus and method for parallel processing according to the present invention performs data-level parallelism or data parallelism on the CTR in consideration of data independence of the CTR.

4 shows a process of the CBC-MAC operation mode.

As shown in FIG. 4, CBC-MAC, a method of ensuring integrity, starts with encrypting a message block like an initial vector (IV), and then uses CBC (Chaining), a chaining technique that uses the value for the next encryption. Cipher Block Chaining). CBC-MAC is a method of using the last encrypted block of the CBC as a MAC, a code for authenticating the integrity of a message. Since the CBC performs chaining to remove independence of block ciphered data, when some data is modulated, the CBC propagates the modulation to the rest of the data. In other words, if the modulation is confirmed using the last block of the CBC as the MAC, the integrity of the data can be verified. As will be described again below, data parallelism is not performed on the CBC-MAC in consideration of the CBC-MAC data dependency.

Figure 5 shows the overall process of the AES-CCM operating mode.

As shown in FIG. 5, AES-CCM can simultaneously guarantee confidentiality and integrity with a single key. In other words, AES-CCM provides message authentication and integrity using the CBC-MAC while providing confidentiality of data using the CTR, which is one of the operation modes. First, CBC is used to create an encryption block to be used as a MAC, and CTR is used to encrypt a message. According to an embodiment, the processing order of the CBC and the CTR may be adjusted.

6 shows an example of an AES-CCM parallel processing method according to the present invention in a flow chart.

In order to utilize data-level parallelism in the message encryption / authentication process (S230), which occupies most of the entire execution time of AES-CCM, there should be no dependency between data. However, unlike the CTR mode, in the case of CBC-MAC, there is a dependency due to chaining between blocks, so a method of parallelizing only CTRs capable of data parallelization can be considered.

1 and 6, first, the core allocator 120 allocates a core of the general-purpose multicore processor 140 to a task to be processed in the AES-CCM (S610). In this case, when there is no dependency between data of the processing target task, the core allocating unit 120 may allocate two or more of the cores of the general purpose multicore process 140 to the processing target task. . On the other hand, when there is a dependency between the data of the processing target task, the core allocating unit 120 may allocate one core among the cores of the general purpose multicore processor 140 to the processing target task. .

In one embodiment, the core allocating unit 120 allocates two or more cores when the processing target task is a CTR (Counter) among the cores of the general-purpose multicore processor 140, and the processing target task. If CBC-MAC (Cipher Block Chaining-Message Authentication Code), one core may be allocated. Of course, if the task to be processed is the CTR, the core allocator 120 may allocate the entire core of the general purpose multicore process.

Next, the data parallelization unit 130 performs data level parallelism or data parallelism on the object to be processed in correspondence with the allocated number of cores (S620). In this case, the data parallelization unit 130 does not perform the data parallelization when there is a dependency between data of the processing target task. That is, the data parallelization unit 130 does not perform the data parallelization when the task to be processed is the CBC-MAC.

Next, the general-purpose multicore processor 140 performs the data-parallelized task to be performed in parallel through the assigned cores in operation S630. In this case, the general purpose multicore processor 140 does not perform parallel processing between tasks. That is, when the processing for the CTR and the CBC-MAC is performed, the processing for the CTR or the CBC-MAC is performed first.

FIG. 7 illustrates a process of the general-purpose multicore processor 140 according to the AES-CCM parallel processing method of FIG. 6.

As illustrated in FIG. 7, the general-purpose multicore processor 140 may be a quad core processor. After processing the data by parallelizing the CTR in all four cores, the CBC- MAC block generation may be performed.

8 is a flowchart showing another example of the AES-CCM parallel processing method according to the present invention.

 When the task parallelization for the CTR and the CBC-MAC and the data parallelization for the CTR are performed simultaneously, the execution time of the AES-CCM can be further reduced.

1 and 8, the AES-CCM parallel processing apparatus 100 according to the present invention further includes the task parallelization unit 110, among the target tasks of the AES-CCM processing, CTR and CBC. Task parallelism for the MAC may be performed (S810).

Then, the core allocating unit 120 allocates a core of the general purpose multicore processor 140 to a processing target task of the general purpose multicore processor 140 AES-CCM (S820). In particular, the core allocator 120 may allocate one core to the CBC-MAC and one or more cores to the CTR among the cores of the general-purpose multicore processor 140. have. Of course, the core allocator 120 may allocate all remaining cores other than the cores allocated to the CBC-MAC to the CTR.

Next, the data parallelization unit 130 performs data parallelization on the object to be processed in correspondence with the allocated number of cores (S830). In particular, the data parallelization unit 130 may perform the data parallelization only for the CTR. As mentioned above, the CBC-MAC has a data dependency.

Next, the general-purpose multicore processor 140 performs the data-parallelized task to be performed in parallel through the allocated cores in operation S840. In particular, the general purpose multicore processor 140 performs the processing for the CTR and the CBC-MAC in parallel.

FIG. 9 illustrates a process of the general-purpose multicore processor 140 according to the AES-CCM parallel processing method of FIG. 8.

As shown in FIG. 9, the general-purpose multicore processor 140 may be a quad core processor, wherein three of the four cores have three cores for the CTR and one core for the CBC-MAC. The allocation to the CTR and the CBC-MAC can be performed in parallel.

On the other hand, the present invention can be embodied as computer readable program codes on a computer readable recording medium. In this case, the program can be programmed through Pthreads (POSIX Threads), which is a standard application programming interface (API) used for writing software that runs in parallel. When the Pthread is used, parallel program compatibility can be provided to operate on various multicore CPUs.

When the present invention is executed through such software, the constituent means of the present invention are code segments for performing necessary tasks. In addition, the program or code segments may be stored in a computer readable medium of a computer or transmitted as a computer data signal coupled with a carrier via a transmission medium or a communication network.

Computer-readable recording media include all kinds of recording devices for storing data that can be read by a computer system. For example, the computer-readable recording medium may include a ROM, a RAM, a CD-ROM, a magnetic tape, a floppy disk, an optical data storage device, and the like. The computer readable recording medium can also be distributed over network coupled computer systems so that the computer can store and execute the code that is readable.

In addition, the present invention can be implemented in a system on chip technology to enable the microprocessor to perform the above-described AES-CCM parallel processing operation. If the present invention is implemented in a microprocessor, the size of various systems can be reduced, the assembly process can be simplified, and manufacturing costs can be reduced.

Hereinafter, the significant effects of the present invention are verified.

10 is a graph showing the results of the AES-CCM parallel processing execution time measurement experiment according to the present invention.

In the experiment, data of various sizes were encrypted / decoded in an experimental environment of an Intel Core2 Quad CPU 2.33GHz. The key length used for the experiment was 128 bits, the initial vector (IV) was 12 bytes, and the tag was 16 bytes. The data was measured from 10MB to 50MB in increments of 10MB and compared with the conventional method using a single core (SEQ) .AES-CCM parallel processing method according to the present invention, a method of performing data parallelization with respect to CTR (CTR ) And the (CTR + CBC) method that simultaneously performs task parallelization and data parallelization for CTR and CBC-MAC.

As shown in FIG. 10, the present invention shows that when data-level parallelism is used only for CTR in AES-CCM, execution time is reduced by about 35% compared to the conventional method. When task-level parallelism and data parallelism are used simultaneously for CBC-MAC, the execution time is reduced by about 50%.

Meanwhile, with the recent increase in semiconductor integration, general purpose multicore CPUs have been installed not only in standalone PCs but also in embedded systems. In this trend, if the parallel processing of the AES-CCM using a general-purpose multi-core CPU according to the present invention, it is possible to propose a way to obtain a cost-effective solution than the existing solutions on the assumption of a dedicated hardware chip It is expected.

As described above, the present invention provides an advantage of improving processing speed and resource efficiency as well as ensuring data confidentiality and integrity simultaneously by parallelizing AES-CCM. In addition, using a general-purpose multicore processor provides the advantage of enabling cost-effective solution development without the need for separate hardware. Furthermore, using Pthreads provides the advantage of being compatible with a variety of CPUs.

So far, the present invention has been described with reference to the embodiments. However, one of ordinary skill in the art will appreciate that the present invention can be implemented in a modified form without departing from the essential technical spirit of the present invention. Therefore, the disclosed embodiments should be considered in an illustrative rather than a restrictive sense. That is, the true technical scope of the present invention is shown in the appended claims, and all differences within the equivalent scope will be construed as being included in the present invention.

1 is a block diagram showing an example of an AES-CCM parallel processing apparatus using a general-purpose multicore process according to the present invention.

2 is a flowchart illustrating a process of AES-CCM.

3 is a diagram illustrating a process of a CTR operation mode;

4 is a diagram illustrating a processing procedure of a CBC-MAC operation mode.

5 is a diagram showing the overall processing of the AES-CCM operating mode.

6 is a flow chart showing an example of an AES-CCM parallel processing method according to the present invention.

7 is a diagram illustrating a processing procedure of a general-purpose multicore processor according to the AES-CCM parallel processing method of FIG. 6.

8 is a flow chart showing another example of the AES-CCM parallel processing method according to the present invention.

9 is a diagram illustrating a processing procedure of a general-purpose multicore processor according to the AES-CCM parallel processing method of FIG. 8.

10 is a graph showing the results of the AES-CCM parallel processing execution time measurement experiment according to the present invention.

Claims (20)

A core allocator configured to allocate one or more cores required according to the processing target task among cores of a general-purpose multicore processor to the processing target task of the AES-CCM; And A data parallelization unit configured to perform data parallelism on the processing target task in response to the number of the assigned cores; The general purpose multicore processor is an AES-CCM parallel processing apparatus using a general-purpose multicore process for performing the data parallelized processing target task through the assigned core in parallel. The method of claim 1, The core allocator may allocate two or more of cores of the general-purpose multicore process to the processing target task when there is no dependency between the data of the processing target task. CCM parallel processing unit. The method of claim 1, If there is a dependency between the data of the task to be processed, the core allocator allocates one core among the cores of the general-purpose multicore process to the processing target task. CCM parallel processing unit. The method of claim 1, The core allocating unit allocates two or more cores when the processing target task is a CTR (Counter) among the cores of the general-purpose multicore process, and the processing target task is a Cipher Block Chaining-Message Authentication Code (CBC-MAC). In the case of AES-CCM parallel processing apparatus using a general-purpose multi-core process, characterized in that allocating one core. 5. The method of claim 4, And the core allocating unit allocates an entire core of the general-purpose multicore process when the task to be processed is the CTR. The method of claim 5, The general-purpose multicore processor, the AES-CCM parallel processing apparatus using a general-purpose multicore processor, characterized in that when performing the processing for the CTR and the CBC-MAC, the first processing for the CTR or the CBC-MAC . The method of claim 1, The data parallelization unit, AES-CCM parallel processing apparatus using a general-purpose multi-core process, if there is a dependency between the data of the task to be processed, the data parallelization. The method of claim 1, The data parallelization unit, AES-CCM parallel processing apparatus using a general-purpose multi-core process, characterized in that the data parallelization does not perform when the task to be processed is CBC-MAC. The method of claim 1, The AES-CCM parallel processing apparatus further includes a task parallelization unit configured to perform task parallelism on CTR and CBC-MAC among tasks to be processed in the AES-CCM. AES-CCM parallel processing unit using. 10. The method of claim 9, The core allocator may allocate one core to the CBC-MAC and one or more cores to the CTR among the cores of the general-purpose multicore process. AES-CCM parallel processing unit. The method of claim 10, The data parallelization unit, AES-CCM parallel processing apparatus using a general-purpose multi-core process, characterized in that for performing the data parallelization only for the CTR. The method of claim 10, The general-purpose multicore processor, the AES-CCM parallel processing apparatus using a general-purpose multicore processor, characterized in that for performing the processing for the CTR and the CBC-MAC in parallel. In a computer system using a general-purpose multicore processor, a core allocation step of allocating one or more cores required according to the processing target task among the cores of the general-purpose multicore processor to the processing target task of AES-CCM; In the computer system, performing data parallelism on the task to be processed in correspondence to the number of allocated cores; And In the general-purpose multi-core processor, AES-CCM parallel processing method using a general-purpose multi-core process comprising the step of processing the data parallelized processing target task through the assigned core in parallel. The method of claim 13, The core allocating step may include allocating two or more cores when the task to be processed is CTR and one core when the task to be processed is CBC-MAC among the cores of the general-purpose multicore process. AES-CCM parallel processing method using a general purpose multicore process. The method of claim 14, And the core allocating step comprises allocating an entire core of the general purpose multicore process when the task to be processed is the CTR. The method of claim 15, Wherein the processing step, AES-CCM parallel processing method using a general-purpose multi-core process, characterized in that when performing the processing for the CTR and the CBC-MAC, the first processing for the CTR or the CBC-MAC . The method of claim 13, The data parallelizing step is the step of performing the data parallelism when the task to be processed is the CBC-MAC, AES-CCM parallel processing method using a general-purpose multi-core process. In a computer system using a general-purpose multicore processor, a task parallelization step of performing task parallelism for CTR and CBC-MAC among the tasks to be processed in the AES-CCM; In the computer system, allocating one core for the CBC-MAC and one or more cores for the CTR among the cores of the general-purpose multicore process; In the computer system, performing data parallelism on the CTR in correspondence with the allocated number of cores; And And in the general-purpose multicore processor, processing the data parallelized CTR and the CBC-MAC in parallel through the allocated cores. A computer-readable recording medium having recorded thereon a program for executing a method according to any one of claims 13 to 18 with a computer. The method of claim 19, The program is a computer-readable recording medium, characterized in that programmed through the Pthread (POSIX Threads), which is a standard application programming interface (API) used for writing software that operates in parallel.

Images