Extended graph connectivity and its gradually increasing parallel complexity
α-connectivity is a graph problem whose complexity gradually increases as the single parameter α grows. It is known that (i) α-connectivity is in NCt when α=c(logn)t−2/2, and (ii) it is P-complete when α=cn e . Unfortunately, the above (i) is only a result on increasing upper bounds. In this paper, we give more reasonable and stronger evidence that the complexity of α-connectivity really increases gradually as α grows. It is shown that α-connectivity can simulate gradually larger-size circuits as α grows. If at most α−1 gates in each level of the circuit can have value 1, then α-connectivity can simulate gradually larger-depth circuits. We finally show evidence suggesting that these simulating powers of α-connectivity are best possible.
Unable to display preview. Download preview PDF.
- [CT91]Cheriyan, J. and Thurimella, R.: Algorithms for parallel k-vertex connectivity and sparse certificates. Proc. 20th ACM Symp. on Theory of Computing (1991) 391–401Google Scholar
- [II93]Iwamoto, C. and Iwama, I.: On Gradually Unparallelizable Graph Problems, Technical Report, 93C-12, Kyushu Univ., Fukuoka (1993)Google Scholar
- [Jo90]Johnson, D. S.: A catalog of complexity classes. Handbook of Theoretical Computer Science (J. van Leeuwen, ed.). Vol. A, 68–161, MIT Press, Amsterdam (1990)Google Scholar
- [KR90]Karp, R. M. and Ramachandran, V.: Parallel algorithms for shared-memory machines. Handbook of Theoretical Computer Science (J. van Leeuwen, ed.). Vol. A, 869–941, MIT Press, Amsterdam (1990)Google Scholar
- [KUW86]Karp, R. M., Upfal, E. and Wigderson, A.: Constructing a perfect matching is in Random NC. Combinatorica 6 1 (1986) 35–48Google Scholar
- [KS91]Khuller, S. and Schieber, B.: Efficient parallel algorithms for testing k-connectivity and finding disjoint s-t paths in graphs. SIAM J. Comput. 20 2 (1991) 352–375Google Scholar