• KIPS Transactions on Computer and Communication Systems
• /
• v.9 no.5
• /
• pp.113-120
• /
• 2020

KISTI(Korea Institute of Science and Technology Information) provides HPC(High Performance Computing) service to users of university, institute, government, affiliated organization, company and so on. The NURION, supercomputer that launched its official service on Jan. 1, 2019, is the fifth supercomputer established by the KISTI. The NURION has 25.7 petaflops computation performance. Understanding how supercomputing services are used and how researchers are using is critical to system operators and managers. It is central to monitor and analysis network traffic. In this paper, we briefly introduce the NURION system and supercomputing service network with security configuration. And we describe the monitoring system that checks the status of supercomputing services in real time. We analyze inbound/outbound traffics and abnormal (attack) IP addresses data that are collected in the NURION supercomputing service network for 11 months (from January to November 1919) using time series and correlation analysis method.

• The KIPS Transactions:PartA
• /
• v.12A no.2 s.92
• /
• pp.171-180
• /
• 2005

In distributed systems data are replicated and stored at several nodes to increase the availability and overall performance. Here Quorum protocol doffing a certain set of replicas required for read/write operation exists for global concurrency control. One of the representative replication Protocols - the Tree Quorum protocol - has a drawback of rapidly growing number of replicas as the level increases, while the Grid protocol requires the same operation cost even without any failure. In this paper, thus, we propose a new replication protocol called hybrid protocol which capitalizes the merits of the existing protocols and solves the problems of them at the same time. The proposed hybrid protocol has very low operation cost in the absence of failure like the tree quorum protocol, and has relatively lower operation cost and higher availability than existing protocols when failure occurs by employing tree architecture as the overall organization while each level of the tree is organized as a row of a grid architecture. It is thus effective to be applied to survival storage system. We conduct cost and availability analysis of the proposed protocol through mathematical modeling, and response time and throughput are compared with those of the Tree Quorum protocol through computer simulation.

• The KIPS Transactions:PartC
• /
• v.9C no.6
• /
• pp.847-856
• /
• 2002

Cellular network applications are growing drastically and this requires a fast and efficient transport method between the base station and the mobile switching center. One possible solution is to use ATM links. The low data rate and small-sized packets in the typical cellular applications imply that significant amount of link bandwidth would be wasted, if this small sized packet is carried by one ATM cell. For efficient operation for such cellular and low bit rate applications, a new type of ATM Adaptation Layer, AAL Type 2, has been proposed. In this paper, the principles of AAL Type 2 are briefly described along with the introduction of other alternatives which have formed the basis for this new AAL. The result from the simulation to study the performance of the AAL Type 2 is discussed from the view point of packet delay and ATM cell use efficiency. Due to the variable size of packets in this application, the fairness issue in serving variable sized packets is also discussed along with the effect of fair queueing algorithm implemented at AAL Type 2.

• The KIPS Transactions:PartD
• /
• v.9D no.4
• /
• pp.571-578
• /
• 2002

In temporal database, extended time dimension for history management brings about complexity of join operation and increased cost. To solve this problem, a method that joins the divided segment time data after partition the time range into fixed time interval is introduced. But existing methods can't solve the ambiguity problem of time border that caused by temporal granularity in the partition point. In this paper, We suggested Fuzzy Minimum Interval Partition (FMIP) method that introduced the possibility distribution of fuzzy theory considered uncertainty time interval border in the partition line.

• The KIPS Transactions:PartC
• /
• v.13C no.7 s.110
• /
• pp.873-880
• /
• 2006

Since various networks are deployed and the most of users request higher mobility, there are many researches about the interworking between widely deployed 3G network and rapidly boarded WLAN. On the other side, WiBro is focused on as a next generation network, because many people expect that WiBro gives satisfaction about the enough mobility and mass data transmission. So the study of the integration between WiBro and cdma2000 will show better effects than the present study of the integration between WLAN and cdma2000. The L2 handoff proposed in this paper takes advantages over the existing L3 handoff scheme because it does not require the L3 procedure for the mobility unlike the L3 handoff. Through extensive computer simulations, the efficiency of the proposed scheme has been validated.

• Journal of Sensor Science and Technology
• /
• v.5 no.3
• /
• pp.65-73
• /
• 1996

One of the important factor in drug stuffs to a patient is to inject exact amount with stable flow rates. Since improper injection amount and flow rates would cause bad effect to recovery of a patient, the detecting sensors with high sensitivity is required for an injection pump systems' performance improvement. In this study, the three sensors, piezo film sensor, photo transistor and photo array, were compared to find best one for an injection pump monitoring system. Using suggested data processing technique and photo array sensors, we could minimize the effect of interference, disturbance, illumination, and sensitivity change caused by sensor's position. According to the experiments, the photo array showed the higher reliance than any other the three types of sensors.

• Journal of Broadcast Engineering
• /
• v.24 no.6
• /
• pp.1122-1133
• /
• 2019

We have designed and implemented a new data processing framework called MOHA(Mtc On HAdoop) which can effectively support Many-Task Computing(MTC) applications in a YARN-based Hadoop platform. MTC applications can be composed of a very large number of computational tasks ranging from hundreds of thousands to millions of tasks, and each MTC application may have different resource usage patterns. Therefore, we have implemented MOHA-TaskExecutor(a pilot-job that executes real MTC application tasks)'s Adaptive Parallel Computability which can adaptively execute multiple tasks simultaneously, in order to improve the parallel computability of a YARN container and the overall system throughput. We have implemented multi-threaded version of TaskExecutor which can "independently and dynamically" adjust the number of concurrently running tasks, and in order to find the optimal number of concurrent tasks, we have employed Hill-Climbing algorithm.

• Journal of Navigation and Port Research
• /
• v.27 no.4
• /
• pp.465-469
• /
• 2003

It has well known that MUSIC and ESPRIT algorithms estimate angle of arrival(AOA) with high resolution by eigenvalue decomposition of the covariance matrix which were obtained from the array antennas, However, the disadvantage of MUSIC and ESPRIT is that they are computationally ineffective, and then they are difficult to implement in real time. the other problem of MUSIC and ESPRIT is to require calibrated antennas with uniform features, and are sensitive ti the manufacturing fault and other physical uncertainties. To overcome these disadvantages, several method using neural model have been study. For multiple signals, those methods require huge training data prior to AOA estimation. This paper proposes the algorithm for AOA estimation by interconnected Hopfield neural model. Computer simulations show the validity of the proposed algorithm. It follows that the proposed method yields better AOA estimates than MUSIC. Moreover, out method does not require huge training procedure and only assigns interconnected coefficients to the neural network prior to AOA estimation.

• Journal of KIISE:Computer Systems and Theory
• /
• v.31 no.11
• /
• pp.635-642
• /
• 2004

The traditional communication protocols such as TCP/IP are not suitable for PC cluster systems because of their high software processing overhead. To eliminate this overhead, industry leaders have defined the Virtual Interface Architecture (VIA). VIA provides two different data transfer mechanisms, a traditional Send/Receive model and the Remote Direct Memory Access (RDMA) model. RDMA is extremely efficient way to reduce software overhead because it can bypass the OS and use the network interface controller (NIC) directly for communication, also bypass the CPU on the remote host. In this paper, we have implemented VIA-based RDMA mechanism in hardware. Compared to the traditional Send/Receive model, the RDMA mechanism improves latency and bandwidth. Our RDMA mechanism can also communicate without using remote CPU cycles. Our experimental results show a minimum latency of 12.5${\mu}\textrm{s}$ and a maximum bandwidth of 95.5MB/s. As a result, our RDMA mechanism allows PC cluster systems to have a high performance communication method.

• KIPS Transactions on Computer and Communication Systems
• /
• v.3 no.6
• /
• pp.179-188
• /
• 2014

Due to the tremendous amount and its rapid increase of network traffic, the performance of network equipments are becoming an important issue. Network filtering is one of primary functions affecting the performance of the network equipment such as a firewall or a load balancer to process the packet. In this paper, we propose a cache based tri method to improve the performance of the existing tri method of searching for network filtering. When several packets are exchanged at a time between a server and a client, the tri method repeats the same search procedure for network filtering. However, the proposed method can avoid unnecessary repetition of search procedure by exploiting cache so that the performance of network filtering can be improved. We performed network filtering experiments for the existing method and the proposed method. Experimental results showed that the proposed method could process more packets up to 790,000 per second than the existing method. When the size of cache list is 11, the proposed method showed the most outstanding performance improvement (18.08%) with respect to memory usage increase (7.75%).