• The KIPS Transactions:PartA
• /
• v.13A no.5 s.102
• /
• pp.387-398
• /
• 2006

In this paper, we propose a new page allocation scheme for flash memory file system. The proposed scheme allocates pages by exploiting the concept of Purity, which is defined as the fraction of blocks where valid Pages and invalid Pages are coexisted. The Pity determines the cost of block cleaning, that is, the portion of pages to be copied and blocks to be erased for block cleaning. To enhance the purity, the scheme classifies hot-modified data and cold-modified data and allocates them into different blocks. The hot/cold classification is based on both static properties such as attribute of data and dynamic properties such as the frequency of modifications. We have implemented the proposed scheme in YAFFS and evaluated its performance on the embedded board equipped with 400MHz XScale CPU, 64MB SDRAM, and 64MB NAND flash memory. Performance measurements have shown that the proposed scheme can reduce block cleaning time by up to 15.4 seconds with an average of 7.8 seconds compared to the typical YAFFS. Also, the enhancement becomes bigger as the utilization of flash memory increases.

• Journal of KIISE:Computer Systems and Theory
• /
• v.32 no.11_12
• /
• pp.557-565
• /
• 2005

This paper presents the development and implementation of a zero-copy file transfer mechanism that improves the efficiency of file transfers for PC cluster systems using hardware-based VIA(Virtual Interface Architecture) network adapters. VIA is one of the representative user-level communication interfaces, but because there is no library for file transfer, one copy occurs between kernel buffer and user boilers. Our mechanism presents a file transfer primitive that does not require the file system to be modified and allows the NIC to transfer data from the kernel buffer to the remote node directly without copying. To do this, we have developed a hardware-based VIA network adapter, which supports the PCI 64bit/66MHz bus and Gigabit Ethernet, as a NIC, and implemented a zero-copy file transfer mechanism. The experimental results show that the overhead of data coy and context switching in the sender is greatly reduced and the CPU utilization of the sender is reduced to $30\%\~40\%$ of the VIA send/receive mechanism. We demonstrate the performance of the zero-copy file transfer mechanism experimentally. and compare the results with those from existing file transfer mechanisms.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.12
• /
• pp.9-16
• /
• 2022

In this paper, we propose a plan to design and utilize the RDBS(Record Block Data file management System) so that data can be recovered when data files in the Combat Management System are mismatched. The CMS(Combat Management System) manages the same files in multiple IPN(Infomation Processing Node) repositories to support multiplexing. However, mismatches in data files can occur due to equipment maintenance or user immaturity. The existing CMS does not manage the history of changes in data files, and when a mismatch occurs, data file were synchronized based on the latest date. But, It is difficult to say that files with the latest date have the highest reliability, and once the file synchronization has progressed, it cannot be recovered with pre-synchronization data. To solve this problem, data was stored and synchronized in units of record blocks using RDBS proposed in this paper, and the Rsync algorithm was used to reduce the overhead of file synchronization due to units of record blocks. SW applied with RDBS was tested for performance in a simulated environment, and it was confirmed that it could be applied to CMS through normal operation confirmation.

• Journal of Korea Society of Industrial Information Systems
• /
• v.18 no.4
• /
• pp.9-18
• /
• 2013

Flash memory is mostly used on smart devices and embedded systems because of its nonvolatile memory, low power, quick I/O, resistant shock, and other benefits. Generally the typical file systems base on the NAND flash memory are YAFFS2, JFFS2, UBIFS, and etc. In this paper, we had variously made an experiment regarding I/O performance using our schemes and the UBIFS of the latest Linux Kernel. The proposed I/O performance analyses were classified as a sequential access and a random access. Our experiment consists of 6 cases using kmalloc(), vmalloc(), and kmem_cache(). As a result of our experiment analyses, the sequential reading and the sequential rewriting increased by 12%, 11% when the Case 2 has applied vmalloc() and kmalloc() to the UBI subsystem and the UBIFS. Also, the performance improved more by 7.82%, 6.90% than the Case 1 at the random read and the random write.

• Proceedings of the Korean Society of Broadcast Engineers Conference
• /
• 2013.11a
• /
• pp.20-23
• /
• 2013

Recently for the transfer of large size video file between sites for high-definition craft editing used by broadcasting company the Public Network is used a lot. In the IP Public Network with QoS(Quality-of-Service) not guaranteed, degradation of performance has several causes. In this paper, we have analyzed the causes of low performance to transfer a large size HD video file in long distance, in order to maximize the effectiveness, the high efficiency network could be implemented enabling the highspeed of HD video file using the connection with networks between hosts through packet creation and multi-session of Parallel TCP. We survey the result of high speed transfer system and verify the efficient transfer method using Public IP Network for large HD video file transfer in broadcasting cooperation.

• Journal of Broadcast Engineering
• /
• v.6 no.3
• /
• pp.246-259
• /
• 2001

Due to recent rapid deployment of Internet streaming service and digital broadcasting service, the issue of how to efficiently support streaming workload in so called "Internet Home Appliance" receives prime interests from industry as well as academia. The underlying dilemma is that it may not be feasible to put cutting edge CPU, boards, disks and other peripherals into that type of device. The primary reason is its cost. Usually, Internet Home Appliances has its dedicated usage, e.g. Internet Radio, and thus it does not require high-end CPU nor high-end Va subsystem. The same reasoning applies to I/O subsystem. In Internet Home Appliances dedicated to handle compressed moving picture, it is not equipped with high end SCSI disk with fast rotational speed. Thus, it is mandatory to devise elaborate software algorithm to exploit the available hardware resources and maximize the efficiency of the system. This paper presents our experiences in the design and implementation of a new multimedia file system which can efficiently deliver the required disk bandwidth for a periodic I/O workload. We have implemented the file system on the Linux operating system, and examined itsperformance under streaming I/O workload. The results of the study show that the proposed file system exhibits superior performance than the Linux Ext2 file system under streaming I/O workload. The result of this work not only contribute to advance the state f art file system technology for multimedia streaming but also put forth the software which is readily available and can be deployed. deployed.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.8 no.5
• /
• pp.1035-1041
• /
• 2004

Currently users are tend to use multimedia data types in their applications. Main features of multimedia data types are large amount of data compared to conventional data types. In this reason, it's hard to load data into main memory and to search. That is the cause of occur disk input and output frequently, and decrease the system performance. In this paper, we describe to have fast and efficient access to multimedia data using index technique. Index method presented by paper consists of two parts : one is index file part for keywords and the other is posting file part for the list of file names. Of course, we use keyword. But user is not charge of memory for the keywords. Users just use natural language to insert, delete and search data what he or she wants. Internally, System makes keywords from natural language to get access to multimedia data. It provides convenience to users. Using this study to develop one's application for multimedia, one may have a chance for advanced performance of a system and getting a result speedily.

• Journal of Information Processing Systems
• /
• v.14 no.5
• /
• pp.1150-1166
• /
• 2018

This paper presents a physical storage design method for image access structures using transformation techniques of multidimensional file organizations in image information systems. Physical storage design is the process of determining the access structures to provide optimal query processing performance for a given set of queries. So far, there has been no such attempt in the image information system. We first show that the number of pages to be accessed decreases as the shape of the given retrieval query region and that of the data page region become similar in the transformed domain space. Using these properties, we propose a method for finding an optimal image access structure by controlling the shapes of the page regions. For the performance evaluation, we have performed many experiments with a multidimensional file organization using transformation techniques. The results indicate that our proposed method is at least one to maximum five times faster than the conventional method according to the query pattern within the scope of the experiments. The result confirms that the proposed physical storage design method is useful in a practical way.

• Journal of Korea Society of Industrial Information Systems
• /
• v.21 no.3
• /
• pp.13-19
• /
• 2016

Spark, an in-memory big-data processing framework is popular to use for real-time processing workload. Spark can store all intermediate data in the cluster memory so that Spark can minimize I/O access. However, when the resident memory of workload is larger that the physical memory amount of the cluster, the total performance can drop dramatically. In this paper, we analyse the factors of bottleneck on PageRank Application that needs many memory through experiment, and cluster the Spark with Tachyon File System for using memory to solve the factor of bottleneck and then we improve the performance about 18%.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.5 no.3
• /
• pp.128-135
• /
• 2010

Recently complexity of embedded software cause to be used real-time operating system (RTOS) to implement various functions in the embedded system. And also, according to requirement of complex functions in embedded systems, the number as well as complexity of tasks get increased continuously. In case that many tasks collaborated in a microprocessor, context switching time between tasks is a overhead waisting a CPU resource. Therefore the time of task context switching is an important factor that affects performance of RTOS. In this paper, we concentrate on the improvement of task context switch for reducing overhead and achieving fast response time in RTOS. To achieve these goal, we suggest multiple register files and task context switching algorithm. By reducing the context switch overhead, we try to ease scheduling and assure fast response times in multitasking environment. As a result, the context switch overhead decreased by 8~16% depend on the number of register files, and some task set which are not schedulable with single register file are schedulable due to that decrease with multiple register files.