• IEIE Transactions on Smart Processing and Computing
• /
• 제2권5호
• /
• pp.282-296
• /
• 2013

With the emergence of cloud computing, more and more sensitive user data are outsourced to remote storage servers. The privacy of users' access pattern to the data should be protected to prevent un-trusted storage servers from inferring users' private information or launching stealthy attacks. Meanwhile, the privacy protection schemes should be efficient as cloud users often use thin client devices to access the data. In this paper, we propose a lightweight scheme to protect the privacy of data access pattern. Comparing with existing state-of-the-art solutions, our scheme incurs less communication and computational overhead, requires significantly less storage space at the user side, while consuming similar storage space at the server. Rigorous proofs and extensive evaluations have been conducted to show that the proposed scheme can hide the data access pattern effectively in the long run after a reasonable number of accesses have been made.

• JSTS:Journal of Semiconductor Technology and Science
• /
• 제14권4호
• /
• pp.391-406
• /
• 2014

The performance of General-Purpose computation on Graphics Processing Units (GPGPU) is heavily dependent on the memory access behavior. This sensitivity is due to a combination of the underlying Massively Parallel Processing (MPP) execution model present on GPUs and the lack of architectural support to handle irregular memory access patterns. Application performance can be significantly improved by applying memory-access-pattern-aware optimizations that can exploit knowledge of the characteristics of each access pattern. In this paper, we present an algorithmic methodology to semi-automatically find the best mapping of memory accesses present in serial loop nest to underlying data-parallel architectures based on a comprehensive static memory access pattern analysis. To that end we present a simple, yet powerful, mathematical model that captures all memory access pattern information present in serial data-parallel loop nests. We then show how this model is used in practice to select the most appropriate memory space for data and to search for an appropriate thread mapping and work group size from a large design space. To evaluate the effectiveness of our methodology, we report on execution speedup using selected benchmark kernels that cover a wide range of memory access patterns commonly found in GPGPU workloads. Our experimental results are reported using the industry standard heterogeneous programming language, OpenCL, targeting the NVIDIA GT200 architecture.

• ETRI Journal
• /
• 제34권1호
• /
• pp.66-75
• /
• 2012

This paper proposes a privacy-preserving database encryption scheme that provides access pattern hiding against a service provider. The proposed scheme uses a session key to permute indices of database records each time they are accessed. The proposed scheme can achieve access pattern hiding in situations in which an adversary cannot access the inside of the database directly, by separating the entity with an index table and data table and permuting both the index and position where the data are stored. Moreover, it is very efficient since only O(1) server computation and communication cost are required in terms of the number of the data stored. It can be applied to cloud computing, where the intermediate entities such as cloud computing service provider can violate the privacy of users or patients.

• 한국경영과학회:학술대회논문집
• /
• 한국경영과학회 1996년도 추계학술대회발표논문집; 고려대학교, 서울; 26 Oct. 1996
• /
• pp.275-278
• /
• 1996

We propose dynamic file allocation method in distributed database management system with changing access patterns. There are a lot of studies on file allocation problem in D-DBMS, and those studies deal with off-line analysis and optimization. Those works are well for systems with static database access patterns, but are inadequate for systems that have changing access patterns. In these systems, dynamic file allocation along with access pattern is more proper. In advance, Brunstrom et al. studied on this area, but they dealt a extremely simplified model. So, we make more practical models to simulate real system. In these models, many factors that were disregard in the advance study are considered. These models are composed with the non-replication system and the replication system. In addition to, we deal with CPU workload balancing in such system in order to improve performance of systems. Our methodology is very simple and realistic, therefore we think that it will give a lot of improvement in D-DBMS with changing access pattern.

• 한국멀티미디어학회논문지
• /
• 제4권6호
• /
• pp.489-499
• /
• 2001

This paper presents the SIC(Size-Interval-Count) prefetching system that can record the file access patterns of applications within a relatively small space of memory based on the repetitiveness of the file access patterns. The SICPS(SIC Prefetching System) is based on knowledge-based prefetching methods which includes high correctness in predicting future accesses of applications. The proposed system then uses the recorded file access patterns, referred to as "SIC access pattern information", to correctly predict the future accesses of the applications. The proposed prefetching system improved the response time by about 40% compared to the general file system and showed remarkable memory efficiency compared to the previously knowledge-based prefetching methods.

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 2006년도 하계종합학술대회
• /
• pp.219-220
• /
• 2006

In this paper, we designed and implemented the Dual Band Dipole Antenna with Tapered Microstrip Balun for WLAN Access Point. Two dipole antennas with different resonant frequency and the antenna structure combined additional line were implemented for dual band performance. In order to feed the balun current, the tapered microstrip balun was used. Produced the Dual Band Antenna shows a special quality. The quality is that all VSWR is less than 1.5 in the 2.4GHz and 5GHz frequency bands in 802.11 standards, and it profits not less than 1.7dBi having typical Dipole Antenna pattern the very "a form of 8"pattern and Omni-directional pattern.

• 한국정보과학회논문지:시스템및이론
• /
• 제28권12호
• /
• pp.674-684
• /
• 2001

본 논문에서는 액세스 패턴의 반복성을 이용하여 비교적 적은 메모리 공간을 사용하는 SIC (SiZe-Interval-Count) 선반입 기법을 제안한다. 최근에 연구되어진 지식기반의 선반입 기법은 응용프로그램의 액세스를 예측하여 정확한 선반입을 수행하는 기법이다. 이들 기법은 응용프로그램의 액세스 패턴을 기록하고, 기록된 액세스 패턴정보를 이용하여 다음에 요청될 블록을 예측하게 된다. 하지만 이 기법은 많은 메모리 공간의 사용을 필요로 한다. 따라서 제안된 선반입 기법에서는 "SIC 액세스 패턴 정보"를 이용하여 반복적인 액세스 패턴을 효율적으로 저장하고, 이를 이용하여 응용프로그램의 다음에 요청될 블록을 정확하게 예측한다. 본 논문의 선반입 기법은 일반 파일시스템에 비해 최고 40%의 응답속도 향상을 가져오며, 기존의 지식기반 선반입 기법에 비해 뛰어난 메모리 효율성을 보여준다.

• 국제학술발표논문집
• /
• The 3th International Conference on Construction Engineering and Project Management
• /
• pp.499-505
• /
• 2009

Along with the changing pattern that the construction project is getting complicated, large-scaled and up-to-dated, necessity of efficient management of manpower and resources of the construction site is being increased. In particular, considering that the construction site requires a lot of labor force and in terms of the construction cost, labor cost shares 30-40%, labor management has become a very important factor. In this study, in order to support a stable and successful construction project implementation and supplement the existing system being utilized at the current site, construction site access management automation system based on hand vascular pattern, a biometrics system, is intended to be developed. Relevant theory for the system development and the problem of existing RFID system were evaluated and based on this findings, a system design and DB composition chart were envisaged. In addition, by applying this developed system to a number of sites, its result was analyzed

• 대한전자공학회:학술대회논문집
• /
• 대한전자공학회 1999년도 추계종합학술대회 논문집
• /
• pp.939-942
• /
• 1999

In this paper, Tele-Test System for ASIC Design is constructed. It consists of the server, and the clients. The server and clients are implemented by Java. Using Java RMI system, the remote access via information network is implemented. In this Tele-test system, fault simulation, test pattern compaction, test pattern generation, and path-delay fault test generation services are implemented. All service can be peformed parallel by network access.

• 한국정보처리학회논문지
• /
• 제4권6호
• /
• pp.1615-1624
• /
• 1997

본 논문에서는 미러링을 이용한 디스크 시스템의 복구기법을 제안하고 분석한다. 디스크 결함발생시 수행되어 지는 복구과정이 사용자 디스크 요구에 대한 평균 응답시간에 미치는 영향을 다룬다. 또한, 복구과정 및 성능에 미치는 응용의 입출력 패턴들의 효과를 분석한다. 실험 결과에 따르면 시스템의 부하가 커질수록 일양접근 패턴에 비해 비일양접근 패턴에 대한 평균 응답시간이 짧아진다. 시뮬레이션 결과는 짧은 지연을 가지는 제안된 복구기법이 기존의 기법보다 더 좋은 성능을 가짐을 보인다.