Вы находитесь на странице: 1из 2

The Dynamics of Grid Computing

December 9, 2011
Given a set N of N nodes i |0 < i < N + 1, i N . Each i has the lowing property: 1. Inertia : Any state i,t at time t is dependant on any previous state i,t1 . This dependancy is expressed with a factor Ti 2. Charisma : It inuences any other k with Charisma Ki 3. It accepts the inuence of any other node with acceptance Ai,k = Ak,i 4. At any time t the node i connects or disconnects from the grid with probability Pc or Pd respectively. 5. Ignore the following for rst run: (a) Target dependant charisma (intensity) . all targets get same (b) Eect of charisma (positive, negative) (c) memory : each stach will depend only on the step right behind it, not any other 6. There is one master node. 7. i,t = Ti (i,t1 ) + vector of A
s s k=i s s

Ai,k Kk (i,t1 ) + Z(t, i) . Z random. Ais the state

8. Fervor : The state of i geting a positive feedback while k is present, and zero or negative otherwise. (opposite: Apathy) TASK: 1. Experimentally determine Pc and Pd 2. Equillibrium conditions and their dependance on dierent known paramemters. 3. state evolution of the master node 4. comparison of evolution of two random nodes 5. breakdown conditons 1

6. Fervor and Apathy condition 7. Find Optimization of Programming Strategy based on 1-6 =Future of NetWorked Computing / Coping with Chaos in Computing Grid Networks

Вам также может понравиться