An algorithm is presented for selecting the k-th smallest element of a totally ordered (but not sorted) set of n elements, 1kn, in the case that a special-purpose sorter is used as a coprocessor. When the pipeline merge sorter is used as the special-purpose sorter, we analyze the comparison complexity of the algorithm for the given capacity of the sorter. The comparison complexity of the algorithm is 1.4167no(n), provided that the capacity of the sorter is 256 elements. The comparison complexity of the algorithm decreases as the capacity of the sorter increases.

A tag-partitioned join algorithm is described. The algorithm partitions only one relation, while other partition-based algorithms partition both relations. It is performed as the joinable tuples of one relation are rearranged and some of them are duplicated according to the original sequence of the join attribute values of the other relation. To do this, the algorithm first finds the positions of all the tuples of the other relation which are joinable with each tuple of one relation, and then partitions joinable tuples of one relation into buckets by using the positions found. Final joining is performed on the partitioned relation and the other relation. We analyze and compare the performance of the algorithm with that of other partition-based join algorithms. The comparison shows that our method is better than other partition-based methods under the practical values of the analysis parameters.

This paper deals with the problem of assigning tasks to the processors of a multiprocessor system such that the sum of execution and communication costs is minimized. If the number of processors is two, this problem can be solved efficiently using the network flow approach pioneered by Stone. This problem is, however, known to be NP-complete in the general case, and thus intractable for systems with a large number of processors. In this paper, we propose a network flow approach for the task assignment problem in homogeneous hypercube networks, i.e., hypercube networks with functionally identical processors. The task assignment problem for an n-dimensional homogeneous hypercube network of N (=2n) processors and M tasks is first transformed into n two-terminal network flow problems, and then solved in time no worse than O(M3 log N) by applying the Goldberg-Tarjan's maximum flow algorithm on each two-terminal network flow problem.

Petri net based satisfiability testing for non-clausal propositional calculus is presented. Every formula in propositional calculus is hierarchically represented by an acyclic free-choice Petri net called logic Petri net. Then a logic Petri net is combined with two way alternating modules, whose number depends upon the number of atoms used in the formula. It is shown that the satisfiability of a formula with non-clausal form can be tested by obtaining reachable firing set of the complete logic Petri net corresponding to the formula under the maximum firing rule.

The use of hypercube multiprocessor computer for solving large-scale linear programming problems with parallel simplex method is presented. The inherent parallelisms involved in each step of the sequential simplex method are investigated and we show how the topological properties of hypercube are effectively applied in the proposed algorithm. The analysis shows that an O(P) speedup with respect to the total running time required by sequential implementations of the simplex method is achieved by using P2p processors of p-dimensional hypercube.

Minimizing intersite data transmissions is important for the join operation between two fragmented relations in distributed databases. In this paper, we examine communication costs of distributed joins when data redundancy and semantic information associated with fragments are not considered. We use a bit vector filtering technique to minimize unnecessary data transmissions in a computer network. The procedure of a distributed join algorithm is proposed and its communication cost is analyzed with filtering error in hashing. We performed computational experiments to show effect of the communication cost of the proposed algorithm in the number of sites and the semijoin selectivities. The experiments also include performance comparison between the proposed algorithm and another semijoin method which recently appeared in the literature. The results show that the communication cost of the proposed algorithm is smaller than that of the semijoin method the fragmented relations.

In this paper, we evaluate the performance of a new join algorithm, called hybrid join, which improves both the sort-merge and the hash-partitioned join algorithms. The hybrid join consists of completely sorting only the smaller relation and partitioning the other one into ranged buckets according to the order statistics of the sorted relation. The final joining is performed on the sorted relation and the ranged buckets. By analytical comparisons, we show that the hybrid join always outperforms the sort-merge join and guarantees better performance than that of the hash-partitioned join in practical situations. The analyses of the performances for the various methods are validated by simulation experiments.

This letter deals with the problem of assigning tasks to the processors of a distributed computing system such that the sum of execution and communication costs is minimized. If the number of processors is two, this problem can be solved efficiently using the network flow approach pioneered by Stone. This problem is, however, known to be NP-complete in the general case, and thus intractable for systems with a large number of processors. Recently, an optimal algorithm with the time complexity of O(n2m3 log n) has been suggested for the problem of assigning m interacting tasks to a linear array of n processors. They solved the problem by using the two-terminal network flow approach. In this letter, we propose a multiterminal network flow approach for the task assignment problem in homogeneous linear array networks. The task assignment problem for a homogeneous linear array network of n processors is first transformed into (n-1) two-terminal network flow problems, and then solved in time no worse than O(nm3) by applying the Goldberg-Tarjan's network flow algorithm for each two-terminal network flow problem.

This paper presents a destributed algorithm that uses weak copy consistency to create mutual exclusion in a distributed computer system. The weak copy consistency is deduced from the uncertainty of state which occurs due to the finite and unpredictable communication delays in a distributed environment. Also the method correlates outdated state information to current state. The average number of messages to enter critical section in the algorithm is n/2 to n messages where n is the number of sites. We show that the algorithm achieves mutual exclusion and the fairness and liveness of the algorithm is proven. We study the performance of the algorithm by simulation technique.

We propose a multistage DeBruijn graph, a generalization of the DeBruijn graph, for dense symmetric interconnection networks in multicomputer systems. The proposed graph is symmetric and contains remarkably large number of vertices than other known symmetric graphs such as the Boolean n-cube and the star graph for given degree and diameter.