boards

本页内容为未名空间相应帖子的节选和存档,一周内的贴子最多显示50字,超过一周显示500字 访问原贴
Science版 - a problem in randomized algorithm
相关主题
The Structure of Scientific Revolutions
sample of cover letter?
problem I never understand since high school
问个非线性科学的问题
Re: Random Walk
A card game
请教一个Monte Carlo摹拟的问题
Re: 关于随机数产生的问题?
Re: 请教:关于fractals and percolation
Re: 把epr实验改进一把
相关话题的讨论汇总
话题: randomized话题: algorithm话题: processors话题: problem话题: broadcast
进入Science版参与讨论
1 (共1页)
s***h
发帖数: 662
1
consider a set of n processors, labeled p1, p2, ...pn. each trying to
broadcast a message. Time moves in discrete steps, and a processor will
succeed in broadcasting its message in step t if and only if no other
processors are trying to broadcast their messages in step t
since the processor cannot communicate directly, they try the following
randomized scheme to resolve contention. For a parameter m to be
determined later, they run a protocol that lasts for m consecutive phases,
each pha
s***h
发帖数: 662
2
ok, got it, it is (e*ln n)*(1+delta)
delta >= sqrt(-ln 0.05/e* ln n)
//relaxed

【在 s***h 的大作中提到】
: consider a set of n processors, labeled p1, p2, ...pn. each trying to
: broadcast a message. Time moves in discrete steps, and a processor will
: succeed in broadcasting its message in step t if and only if no other
: processors are trying to broadcast their messages in step t
: since the processor cannot communicate directly, they try the following
: randomized scheme to resolve contention. For a parameter m to be
: determined later, they run a protocol that lasts for m consecutive phases,
: each pha

1 (共1页)
进入Science版参与讨论
相关主题
Re: 把epr实验改进一把
Re: 来来来, 讨论randomness
一个关于random close pack的问题 (转载)
BH误发2个a7RIII给一老外,白得3200刀的机身后,他。。。
疫苗公司的文件,大家来分析分析 (转载)
超难概率题 (转载)
这样还能算Randomized sample吗
黨又贏了:中國封鎖諾貝爾獎新聞(紐約時報)
US 'Serious Considering' Selling Taiwan F16C/Ds
黨又贏了:中國封鎖諾貝爾獎新聞(紐約時報) (转载)
相关话题的讨论汇总
话题: randomized话题: algorithm话题: processors话题: problem话题: broadcast