Deterministic distributed algorithms for listing cliques of size p in n1 - 2/p + o(1) rounds in the Congest model are given and are the first non-trivial deterministic distributedgorithms to be given.
This paper introduces the optimizations included in mDRACO, such as map-based association of pheromone with arbitrary-length component sequences and relative path storage, and gives promising results on the Costas-array problem.
A main step of the algorithm, called ancestral reconstruction, performs accurately in an idealized setting with no inbreeding but performs poorly in random mating populations, and a Belief-Propagation based heuristic is introduced that accounts for the inbreeding and performs much better in simulations.