• KIPS Transactions on Software and Data Engineering
• /
• v.7 no.5
• /
• pp.177-188
• /
• 2018

The scale of graph data has been increased rapidly because of the growth of mobile Internet applications and the proliferation of social network services. This brings upon the imminent necessity of efficient distributed and parallel graph processing approach since the size of these large-scale graphs are easily over a capacity of a single machine. Currently, there are two popular parallel graph processing approaches, vertex-centric graph processing and block centric processing. While a vertex-centric graph processing approach can easily be applied to the parallel processing system, a block-centric graph processing approach is proposed to compensate the drawbacks of the vertex-centric approach. In these systems, the initial quality of graph partition affects to the overall performance significantly. However, it is a very difficult problem to divide the graph into optimal states at the initial phase. Thus, several dynamic load balancing techniques have been studied that suggest the progressive partitioning during the graph processing time. In this paper, we present a load balancing algorithms for the block-centric graph processing approach where most of dynamic load balancing techniques are focused on vertex-centric systems. Our proposed algorithm focus on an improvement of the graph partition quality by dynamically reassigning blocks in runtime, and suggests block split strategy for escaping local optimum solution.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.13 no.1
• /
• pp.63-73
• /
• 1988

In this paper, an algorithm is proposed to evaluate the source-to-terminal reliability, the probability that a source node can communicate with a terminal node, in a probabilistic derected graph. By using Satyanaratana's factoring $theorem^{(7)}$, the original graph can be partitioned into two reduced graphs obtained by contracting and deleting the edge connected to the source node in the probabilistic directed graph. The edge removal proposed in this paper and the general series-parallel reduction can then be applied to the reduced graph. This edge reduction can be applied recursively to the reduced graphs until a source node can be connected to a terminal node by one edge. A computer program which can be applied to evaluating the source-to-terminal reliability in a complex and large network has also been developed.

• Transactions of the Korean Society of Mechanical Engineers A
• /
• v.32 no.9
• /
• pp.761-769
• /
• 2008

The majority of mechanical parts are manufactured by milling machines. Hence, geometrically efficient algorithms for tool path generation and physical considerations for better machining productivity with guarantee of machining safety are the most important issues in milling tasks. In this paper, an optimized path generation algorithm for direction parallel milling which is commonly used in the roughing stage is presented. First of all, a geometrically efficient tool path generation algorithm using an intersection points-graph is introduced. Although the direction parallel tool path obtained from geometric information have been successful to make desirable shape, it seldom consider physical process concerns like cutting forces and chatters. In order to cope with these problems, an optimized tool path, which maintains constant MRR in order to achieve constant cutting forces and to avoid chatter vibrations at all time, is introduced and the result is verified. Additional tool path segments are appended to the basic tool path by using a pixel based simulation technique. The algorithm has been implemented for two dimensional contiguous end milling operations, and cutting tests are conducted by measuring spindle current, which reflects machining situations, to verify the significance of the proposed method.

• Journal of the Korean Institute of Telematics and Electronics C
• /
• v.35C no.6
• /
• pp.20-29
• /
• 1998

This paper proposes the algorithms that design the highly parallel multiple-valued logic curcuit and assign the code to each node of DCG(Directed Cyclic Graph) of length 1. The conventional Nakajima's algorithm have some problems, so this paper introduce the matrix equation from DCG of length 1 and proposes circuit design algorithms according to the DCG of length 1. Using the proposed circuit design algorithms in this paper, it become realized that was not able to design from Nakajima's algorithm. Also, making a comparision between the circuit design using Nakajima's algorithm and this paper's, we testify that proposed paper's algorithm is able to realize more optimized circuit design. According to proposed curcuit design algorithm in this paper, it is possible to design curcuit that DCG have natural number, so it have the following advantages; reduction of the curcuit input/output digits, simplification of curcuit composition, reduction of computation time and cost. And we show compatibility and verification about this paper's algorithm.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.7
• /
• pp.2060-2071
• /
• 2000

Scheduling parallel tasks, represented as a Directed Acyclic Graph (DAG) or task graph, on a multiprocessor system has been an important research area in the past decades. List scheduling algorithms assign priorities to a node or an edge in an input DAG, and then generate a schedule according to the assigned priorities. This appear proposes a list scheduling algorithms with effective method of priority assignments. The paper also analyzes the worst case performance and optimality condition for the proposed algorithm. The performance comparison study shows that the proposed algorithms outperforms existing scheduling algorithms especially for input DAGs with high communication overheads. The performance improvement over existing algorithms becomes larger as the input DAG becomes more dense and the level of parallelism in the DAG is increased.

• Journal of KIISE:Computer Systems and Theory
• /
• v.35 no.6
• /
• pp.263-272
• /
• 2008

We propose and analyze a new interconnection network, called petersen-torus(PT) network based on well-known petersen graph. PT network has a smaller diameter and a smaller network cost than honeycomb torus with same number of nodes. In this paper, we propose optimal routing algorithm and hamiltonian cycle algorithm. We derive diameter, network cost and bisection width.

• ETRI Journal
• /
• v.20 no.4
• /
• pp.327-345
• /
• 1998

This paper deals with the problem of one-to-one mapping of 2$^n$ task modules of a parallel program to an n-dimensional hypercube multicomputer so as to minimize the total communication cost during the execution of the task. The problem of finding an optimal mapping has been proven to be NP-complete. First we show that the mapping problem in a hypercube multicomputer can be transformed into the problem of finding a set of maximum cutsets on a given task graph using a graph modification technique. Then we propose a repeated mapping scheme, using an existing graph bipartitioning algorithm, for the effective mapping of task modules onto the processors of a hypercube multicomputer. The repeated mapping scheme is shown to be highly effective on a number of test task graphs; it increasingly outperforms the greedy and recursive mapping algorithms as the number of processors increases. Our repeated mapping scheme is shown to be very effective for regular graphs, such as hypercube-isomorphic or 'almost' isomorphic graphs and meshes; it finds optimal mappings on almost all the regular task graphs considered.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.33B no.9
• /
• pp.49-61
• /
• 1996

Back-propagation(BP) algorithm needs a lot of time to train the artificial neural network (ANN) to get high accuracy level in classification tasks. So there have been extensive researches to process back-propagation algorithm on parallel processors. This paper prsents a linear systolic array which calculates forward-backward propagation of BP algorithm at the same time using effective space-time transformation and PE structure. First, we analyze data flow of forwared and backward propagations and then, represent the BP algorithm into data dapendency graph (DG) which shows parallelism inherent in the BP algorithm. Next, apply space-time transformation on the DG of ANN is turn with orthogonal direction projection. By doing so, we can get a snakelike systolic array. Also we calculate the interval of input for parallel processing, calculate the indices to make the right datas be used at the right PE when forward and bvackward propagations are processed in the same PE. And then verify the correctness of output when forward and backward propagations are executed at the same time. By doing so, the proposed system maximizes parallelism of BP algorithm, minimizes th enumber of PEs. And it reduces the execution time by 2 times through making idle PEs participate in forward-backward propagation at the same time.

• Journal of KIISE:Computer Systems and Theory
• /
• v.26 no.10
• /
• pp.1258-1263
• /
• 1999

본 논문에서는 CRCW(Concurrent Read Concurrent Write)와 CREW(Concurrent Read Exclusive Write) PRAM(Parallel Random Access Machine) 모델에서 무방향성 그래프 G=(V, E)의 극대 매칭을 구하기 위해 간결한 랜덤 병렬 알고리즘을 제안한다. CRCW PRAM 모델에서 m개의 선을 가진 그래프에 대해, 제안된 매칭 알고리즘은 m개의 프로세서 상에서 {{{{ OMICRON (log m)의 기대 수행 시간을 가진다. 또한 CRCW 알고리즘을 CREW PRAM 모델에서 구현한 CREW 알고리즘은 OMICRON (log^2 m)의 기대 수행 시간을 가지지만,OMICRON (m/logm) 개의 프로세서만을 가지고 수행될 수 있다.Abstract This paper presents simple randomized parallel algorithms for finding a maximal matching in an undirected graph G=(V, E) for the CRCW and CREW PRAM models. The algorithm for the CRCW model has {{{{ OMICRON (log m) expected running time using m processors, where m is the number of edges in G We also show that the CRCW algorithm can be implemented on a CREW PRAM. The CREW algorithm runs in {{{{ OMICRON (log^2 m) expected time, but it requires only OMICRON (m / log m) processors.

• Journal of KIISE:Computer Systems and Theory
• /
• v.29 no.2
• /
• pp.65-74
• /
• 2002

Recently star networks are considered as attractive alternatives to the widely used hypercube for interconnection networks in parallel processing systems by many researchers. One of the fundamental communication problems on star graph networks is broadcasing In this paper we consider the broadcasting problems in star graph networks using wormhole routing. In wormhole routed system minimizing link contention is more critical for the system performance than the distance between two communicating nodes. We use Hamiltonian paths in star graph to set up link-disjoint communication paths We present a broadcast algorithm in n-dimensional star graph of N(=n!) nodes such that the total completion time is no larger than $([long_n n!]+1)$ steps where $([long_n n!]+1)$ is the lower bound This result is significant improvement over the previous n-1 step broadcasting algorithm.