作 者:
;
;
机构地区:
中国科学技术大学计算机科学与技术学院
出 处:
《计算机学报》
1990年第2期88-100,共13页
摘 要:
本文通过将递归网络E′(m,n)按树形展开,应用组合计数方法导出了(m,n)选择网络(1≤mm)和(当n》m)。
Through tree representation of the recursive network E(m, n) and the combinatorial enumeration, a new upper bound of delay time in the (m,n) selection network (1≤m≤n) is derived which is logn+logm log log (n/m)+ 0(log2mloglogm + log mlogloglog n/m) when n > 4m, log n+log mloglog(n/m) + O(logloglogn) when n》m, and O (log2m) when 2m≤n≤4m. It is an improvement over A. C. Yao's best result
关 键 词:
选择网络
延迟时间
领 域:
[自动化与计算机技术]
[自动化与计算机技术]