• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.2
• /
• pp.85-91
• /
• 2014

Cloud storage system uses a distributed file system for storing and managing data. Traditional distributed file system makes a triplication of data in order to restore data loss in disk failure. However, enforcing data replication method increases storage utilization and causes extra I/O operations during replication process. In this paper, we propose a data replication method using erasure codes in cloud storage system to improve storage space efficiency and I/O performance. In particular, according to data access frequency, the proposed method can reduce the number of data replications but using erasure codes can keep the same data recovery performance. Experimental results show that proposed method improves performance in storage efficiency 40%, read throughput 11%, write throughput 10% better than HDFS does.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.9 no.5
• /
• pp.129-132
• /
• 2009

Frame erasure is one of the most difficult problems in voice over IP (VoIP) networks and is a major source of speech quality degradation. In this paper, a frame erasure concealment algorithm based on nonlinear regression analysis is presented to minimize speech quality deterioration in code-excited linear prediction (CELP) based coders. We applied the proposed scheme to the ITU-T G.729 standard and obtained improved perceptual evaluation of speech quality (PESQ) scores compared to the conventional methods.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.29A no.11
• /
• pp.1-13
• /
• 1992

In this DQDB networks, erasure nodes may be allocated on the bus and release slots which already had been read by destination nodes. An important function of the erasure node must be to eliminate as many requests as erased slots used by downstream nodes. In the networks with priorities, following two additional issues must be solved : (1) accurately to guess for which priority the erased slots is to be used, and (2) to find in a short period a request of that priority to cancel. In this paper, we propose two erasure node algorithms based on elimination of the requests. In these algorithms, the requests are eliminated excessively at low load but optimum elimination can be achieved at full load. Simulation results show that these two algorithms have similar access delay to that of existing algorithms at low load and better performance at high load.

• Proceedings of the KSPS conference
• /
• 2004.05a
• /
• pp.235-238
• /
• 2004

Frame erasures cause speech quality degradation in wireless communication networks or packet networks. The degradation becomes worse when consecutive frame erasures occur. Speech coders have a frame erasure concealment(FEC) mechanism to compensate for frame erasures. It is meaningful to evaluate the performance of FEC mechanisms for frame erasures that occur in communications networks. In this paper, various frame erasures are designed. And the FEC algorithms of speech coders are evaluated and analyzed with the Perceptual Evaluation of Speech Quality(PESQ). It is found that the performances vary in accordance with frame erasure types, frame erasure rates, and utterance lengths.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.30A no.5
• /
• pp.1-15
• /
• 1993

In DQDB networks, the bandwidth can be increased considerably be using the EN(Erasure Node) algorithms and DR(Destination Release) algorithms. However, the important issue in implementing them is using method of extra capacity fairly. To improve it, this paper proposes AEN(Adaptive Erasure Node) algorithm which erasure function is activated by network traffic load. Its functional architecture consists of SESM, RCSM, LMSM in addition to the basic DQDB state machines (DQSM, RQM). The SESM and RCSM state machines are placed in front of the DQSM and RQM state machines in order for the node to take advantage of the newly cleared slots. This paper also presents some simulation results showing the effect of AEN algorithm on access delay, throughput and segment erasing ratio in the single and multiple priority networks. The results show that the AEN algorithm offer the better performance characteristics than existing algorithms under overload conditions.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.10 no.11
• /
• pp.5290-5304
• /
• 2016

With the growth of the massive amount of data, the failure probability of the cloud storage node is becoming more and more big. A single fault tolerance strategy, such as replication and erasure codes, has some unavoidable disadvantages, which can not meet the needs of the today's fault tolerance. Therefore, according to the file access frequency and size, an adaptive hybrid redundant fault tolerance strategy is proposed, which can dynamically change between the replication scheme and erasure codes scheme throughout the lifecycle. The experimental results show that the proposed scheme can not only save the storage space(reduced by 32% compared with replication), but also ensure the fast recovery of the node failures(increased by 42% compared with erasure codes).

• Journal of Communications and Networks
• /
• v.10 no.4
• /
• pp.384-395
• /
• 2008

Random network coding can be viewed as a single block code applied to all source packets. To manage the concomitant high coding complexity, source packets can be partitioned into generations; block coding is then performed on each set. To reach a better performance-complexity tradeoff, we propose a novel concatenated network code which mixes generations while retaining the desirable properties of generation-based coding. Focusing on the code's erasure performance, we show that the probability of successfully decoding a generation on erasure channels can increase substantially for any erasure rate. Using both analysis (for small networks) and simulations (for larger networks), we show how the code's parameters can be tuned to extract best performance. As a result, the probability of failing to decode a generation is reduced by nearly one order of magnitude.

• Proceedings of the Korean Information Science Society Conference
• /
• 2010.06b
• /
• pp.347-351
• /
• 2010

SSD의 칩을 구성하는 셀의 집적도가 높아지고 시스템 오류가 높아짐에 따라, 저장장치의 고신뢰성을 보장해야 하는 필요성은 점점 증가하고 있다. 현재 SSD 의 오류를 복구하기 위한 방법으로 ECC 가 사용되고 있으나 비트단위의 오류복구만 가능하여 복구 성능의 한계를 가지고 있다. 본 논문에서는 SSD 계층구조인 하나의 칩을 디스크로 설정하고 내부에 RAID 시스템을 구성하여 칩 단위의 오류 복구 기법을 제안한다. 제안된 칩 단위의 오류 복구 기법은 기존 하드디스크 기반 RAID 시스템에 적용 되었던 세 가지 Erasure 코드들을 SSD 에 적용한 방법이다. 또한 실험을 통해 각각의 Erasure 코드 성능을 비교하였다.

• Journal of KIISE:Information Networking
• /
• v.31 no.1
• /
• pp.1-10
• /
• 2004

This study discusses Application-layer FEC using erasure codes. Because of the simple decoding process, erasure codes are used effectively in Application-layer FEC to deal with Packet-level errors. The large number of parity packets makes the loss rate to be small, but causes the network congestion to be worse. Thus, a redundancy control algorithm that can adjust the number of parity packets depending on network conditions is necessary. In addition, it is natural that high-priority frames such as I frames should produce more parity packets than low-priority frames such as P and B frames. In this paper, we propose a redundancy control algorithm that can adjust the amount of redundancy depending on the network conditions and depending on data priority, and test the performance in simple links and congestion links.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.47 no.6
• /
• pp.47-56
• /
• 2010

HDD-based RAIDs have been used in high-capacity storage systems for traditional data server. However, their data reliability are relatively low and they consume lots of power since hard disk drive is weak on shock and its power consumption is high due to frequent spindle motor operation. Therefore, this paper presents new SSD based RAID system architecture using various erasure codes. The proposed methode applys Reed-Solomon, EVENODD, and Liberation coding schemes onto file system level and device driver level, respectively. Besides, it uses data allocation method to minimize the side effect of reducing the lifespan of SSD. Detail experimental results show that Liberation code increase wear-leveling rates of SSD based RAID-6 more than other codes. The SSD based RAID system applying erasure codes at the device driver level shows better performance than that at the file system level. I/O performance of RAID-6 system using SSD is 4.5%~8.5% higher than that of using HDD and the power consumption of the RAID system using SSD is 18%~40% less than that of using HDD.