Distributed Algorithms for Connectivity and MST in Large Graphs with Efficient Local Computation
Eric Ajieren, Khalid Hourani, William K. Moses, Gopal Pandurangan · 2022
We study distributed algorithms for large-scale graphs, focusing on the fundamental problems of connectivity and minimum spanning tree (MST). We consider the k-machine model, a well-studied model for distributed computing for large-scale graph computations, where k ≥ 2 machines jointly perform computations on graphs with n nodes (typically, n ≫ k). The input graph is assumed to be initially randomly partitioned among the k machines, a common implementation in many real-world systems. Communication is point-to-point, and the goal is to minimize the number of communication rounds (denoted Tc) of the computation.