Coherent randomized benchmarking

Abstract: Randomized benchmarking is a powerful technique to efficiently estimate the performance and reliability of quantum gates, circuits and devices. Here we propose to perform randomized benchmarking in a coherent way, where superpositions of different random sequences rather than independent samples are used. We show that this leads to a uniform and simple protocol with significant advantages with respect to gates that can be benchmarked, and in terms of efficiency and scalability. We show that e.g. universal gate sets, the set of $n-$qudit Pauli operators or more general sets including arbitrary unitaries, as well as a particular $n-$qudit Clifford gate using only the Pauli set, can be efficiently benchmarked. The price to pay is an additional complexity to add control to the involved quantum operations. However we demonstrate that this can be done by using auxiliary degrees of freedom that are naturally available in basically any physical realization, and are independent of the gates to be tested.