Search by Lackadaisical Quantum Walk with Symmetry Breaking

Abstract

The lackadaisical quantum walk is a lazy version of a discrete-time, coined quantum walk, where each vertex has a weighted self-loop that permits the walker to stay put. They have been used to speed up spatial search on a variety of graphs, including periodic lattices, strongly regular graphs, Johnson graphs, and the hypercube. In these prior works, the weights of the self-loops preserved the symmetries of the graphs. In this paper, we show that the self-loops can break all the symmetries of vertex-transitive graphs while providing the same computational speedups. For vertex-transitive graphs with large numbers of vertices, only the weight of the self-loop at the marked vertex matters, and the remaining self-loop weights can be chosen randomly.