Performance measures for distributed search Academic Article uri icon


  • The waiting periods were probably the trigger for the first discussion in the literature of how one can measure and compare the computational efforts of DisCSP algorithms. Yokoo [6 2, 63] proposes to divide the time that the algorithm consumes into rounds, where at the start of each round at least one agent has the necessary information to carry on its computation. The division of distributed computation into strict rounds is a simplification of the dependency between computations of different agents. An example in Section 10.3 demonstrates that …

publication date

  • January 1, 2008