Discrete optimisation problems on an adiabatic quantum computer

Abstract

In the recent years the field of adiabatic quantum computing has gained importance due to the advances in the realisation of such machines, especially by the company D-Wave Systems. In contrast to a quantum computer in the conventional sense, an adiabatic quantum computer can solve a discrete optimisation problem by encoding the objective function into a quantum mechanical system. By carefully evolving the system in time from an initial state into the lowest energy state, the optimisation problem is solved. Due to the quantum nature of the device it is assumed that there is a substantial speedup compared to classical HPC facilities.

The D-Wave machines are capable of finding the minimum of a subclass of the Ising model, a quadratic unconstrained binary optimisation problem. In this paper, we present a way of mapping more general problems to the subclass of Ising problems in order to make them solvable on an adiabatic quantum computer. As an example, we choose the maximum clique problem where one needs to find the maximal complete subgraph in an undirected graph. The corresponding decision problem is NP-complete. It has applications in social media or computational chemistry. Moreover, the solution of the maximum clique problem can be used to increase the number of logical variables which can be realised on a D-Wave machine. In addition, we investigate the applicability of adiabatic quantum computing to the scheduling optimisation of satellite missions performed by the German Aerospace Center.