Quantum Simulations of Classical Random Walks and Undirected Graph Connectivity

While it is straightforward to simulate a very general class of random processes space-efficiently by non-unitary quantum computations (e.g., quantum computations that allow intermediate measurements to occur), it is not currently known to what extent restricting quantum computations to be unitary affects the space required for such simulations. This paper presents a method by which a limited class of random processes—random walks on undirected graphs—can be simulated by unitary quantum computations in a space-efficient (and time-efficient) manner. By means of such simulations, it is demonstrated that the undirected graph connectivity problem for regular graphs can be solved by one-sided error quantum Turing machines that run in logspace and require a single measurement at the end of their computations. It follows that symmetric logspace is contained in a quantum analogue of randomized logspace that disallows intermediate measurements.