• Journal of KIISE:Information Networking
• /
• v.29 no.5
• /
• pp.466-472
• /
• 2002

We investigate the load sharing problem between the umbrella cell and the overlaid tells in hierarchical cell structure. A load sharing strategy is proposed and its performance is evaluated by simulation when it is used for transmission of the mobile Internet traffic using the High Speed Downlink Packet Access scheme. The results show that, with the proposed strategy, the microcell backs well the overlaid picocells up, especially when a specific picocell cluster suffers unusual heavy load condition. By using the strategy, we can reduce the installation cost, otherwise needed for increasing the system capacity of every picocell cluster to cope with the unusual heavy load.

• The Transactions of the Korean Institute of Power Electronics
• /
• v.9 no.4
• /
• pp.341-349
• /
• 2004

A parallel operation of Uninterruptible Power Supply(UPS) systems is used to increase power capacity of the system or to secure higher reliability at critical loads. In the conventional parallel operation, the load-sharing control to maintain the current balance is the most important, since the load-sharing is very sensitive to discord between components of each module, amplitude/phase difference, line impedance, output LC filter, and so on. To solve these problems various control algorithms are researching. However, these methods cannot apply to the different ratings of UPS. In the case, master and slave control algorithm for parallel operation is adequate. However, if the UPS ratings are different, the value of passive filters L, C is different, and it affects the sharing of current. This paper presents general problems of conventional parallel operation systems, and control strategy for parallel operation with different ratings. The validity of the proposed control strategy is investigated through simulation and experiment in the parallel operation system with two 3-phase UPS systems.

• The KIPS Transactions:PartA
• /
• v.10A no.4
• /
• pp.331-338
• /
• 2003

Cluster is the cost-effective parallel processing environment, and consists of the off-the-shelf computers connected by the computer net works. The characteristics of cluster are the node heterogeneity, the variety of node load, and the variety of network load. Because these characteristics influence the performance of parallel program executions, the load sharing for cluster is important, and by using the proper load sharing strategy, we can reduce the execution time of parallel programs. In this paper, we propose modified GSS algorithm, αGSS. In the proposed load sharing algorithms α GSS, the size of tasks are decided using the BogoMIPS of node. From the result of out experiments, we conclude that the proposed αGSS algorithm is effective in the heterogeneous cluster.

• Journal of Korea Multimedia Society
• /
• v.7 no.2
• /
• pp.264-271
• /
• 2004

A load-sharing algorithm must deal with load imbalance caused by characteristics of a network and heterogeneity of nodes in Internet-based clustering systems. This paper has proposed the Efficient Load-Sharing algorithm. Efficient-Load-Sharing algorithm creates a scheduler based on the WF(Weighted Factoring) algorithm and then allocates tasks by an adaptive granularity strategy and the refined fixed granularity algorithm for better performance. In this paper, adaptive granularity strategy is that master node allocates tasks of relatively slower node to faster node and refined fixed granularity algorithm is to overlap between the time spent by slave nodes on computation and the time spent for network communication. For the simulation, the matrix multiplication using PVM is performed on the heterogeneous clustering environment which consists of two different networks. Compared to other algorithms such as Send, GSS and Weighted Factoring, the proposed algorithm results in an improvement of performance by 75%, 79% and 17%, respectively.

• The Transactions of the Korea Information Processing Society
• /
• v.6 no.12
• /
• pp.3456-3468
• /
• 1999

Most of previous parallel join algorithms assume a database partition system(DPS), where each database partition is owned by a single processing node. While the DPS is novel in the sense that it can interconnect a large number of nodes and support a geographically distributed environment, it may suffer from poor facility for load balancing and system availability compared to the database sharing system(DSS). In this paper, we propose a dynamic load balancing strategy by exploiting the characteristics of the DSS, and then extend the conventional hash join algorithms to the DSS by using the dynamic load balancing strategy. With simulation studies under a wide variety of system configurations and database workloads, we analyze the effects of the dynamic load balancing strategy and differences in the performances of hash join algorithms in the DSS.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2002.11a
• /
• pp.405-408
• /
• 2002

Cluster is the cost-effective parallel processing environment, and consists of the off-the-shelf computers connected by the computer networks. The characteristics of cluster are the node heterogeneity, the variety of node load, and the variety of network load. Because these characteristics influence the performance of parallel program executions, the load sharing for cluster is important, and by using the proper load sharing strategy, we can reduce the execution time of parallel programs. In this paper, we propose modified GSS algorithm, aGSS. In the proposed load sharing algorithms aGSS, the size of tasks are decided using the BogoMTPS of node. From the result of out experiments, we conclude that the proposed aGSS algorithm is effective in the heterogeneous cluster.

• The KIPS Transactions:PartA
• /
• v.8A no.2
• /
• pp.117-124
• /
• 2001

At NOW (Networks Of Workstations), the load sharing is very important role for improving the performance. The known load sharing strategy is fixed-granularity, variable-granularity and adaptive-granularity. The variable-granularity algorithm is sensitive to the various parameters. But Send algorithm, which implements the fixed-granularity strategy, is robust to task granularity. And the performance difference between Send and variable-granularity algorithm is not substantial. But, in Send algorithm, the computing time and the communication time are not overlapped. Therefore, long latency time at the network has influence on the execution time of the parallel program. In this paper, we propose the preSend algorithm. In the preSend algorithm, the master node can send the data to the slave nodes in advance without the waiting for partial results from the slaves. As the master node sent the next data to the slaves in advance, the slave nodes can process the data without the idle time. As stated above, the preSend algorithm can overlap the computing time and the communication time. Therefore we reduce the influence of the long latency time at the network and the execution time of the parallel program on the NOW. To compare the execution time of two algorithms, we use the $320{\times}320$ matrix multiplication. The comparison results of execution times show that the preSend algorithm has the shorter execution time than the Send algorithm.

• The KIPS Transactions:PartA
• /
• v.10A no.3
• /
• pp.215-224
• /
• 2003

Since there are various networks and heterogeneity of nodes in Internet, the existing load-sharing algorithms are hardly adapted for use in Internet-based clustering systems. Therefore, in Internet-based clustering systems, a load-sharing algorithm must consider various conditions such as heterogeneity of nodes, characteristics of a network and imbalance of load, and so on. This paper has proposed an expanded-WF algorithm which is based on a WF (Weighted Factoring) algorithm for load-sharing in Internet-based clustering systems. The proposed algorithm uses an adaptive granularity strategy for load-sharing and duplicate execution of partial job for fault-tolerance. For the simulation, the to matrix multiplication using PVM is performed on the heterogeneous clustering environment which consists of two different networks. Compared to other algorithms such as Send, GSS and Weighted Factoring, the proposed algorithm results in an improvement of performance by 55%, 63% and 20%, respectively. Also, this paper shows that It can process the fault-tolerance.

• Journal of Power Electronics
• /
• v.10 no.6
• /
• pp.673-679
• /
• 2010

Input-series-output-parallel (ISOP) connected DC-DC converters enable low voltage rating switches to be used in high voltage input applications. In this paper, a DSP is adopted to generate digital phase-shifted PWM signals and to fulfill the closed-loop control function for ISOP connected two full-bridge DC-DC converters. Moreover, a stable output current sharing control strategy is proposed for the system, with which equal sharing of the input voltage and the load current can be achieved without any input voltage control loops. Based on small signal analysis with the state space average method, a loop gain design with the proposed scheme is made. Compared with the conventional IVS scheme, the proposed strategy leads to simplification of the output voltage regulator design and better static and dynamic responses. The effectiveness of the proposed control strategy is verified by the simulation and experimental results of an ISOP system made up of two full-bridge DC-DC converters.

• Journal of Power Electronics
• /
• v.16 no.6
• /
• pp.2315-2326
• /
• 2016

This paper discusses the elimination of DC voltage deviation and the enhancement of load current sharing accuracy in multi-terminal high voltage direct current (MT-HVDC) systems. In order to minimize the power line losses in different parallel network topologies and to insure the stable operation of systems, a decentralized control method based on a modified droop control is presented in this paper. Averaging the DC output voltage and averaging the output current of two neighboring converters are employed to reduce the congestion of the communication network in a control system, and the decentralized control method is implemented. By minimizing the power loss of the cable, the optimal load current sharing proportion is derived in order to achieve rational current sharing among different converters. The validity of the proposed method using a low bandwidth communication (LBC) network for different topologies is verified. The influence of the parameters of the power cable on the control system stability is analyzed in detail. Finally, transient response simulations and experiments are performed to demonstrate the feasibility of the proposed control strategy for a MT-HVDC system.