Quantum Advantage of Noisy Grover's Algorithm

Abstract: Quantum advantage is the core of quantum computing. Grover's search algorithm is the only quantum algorithm with proven advantage to any possible classical search algorithm. However, realizing this quantum advantage in practice is quite challenging since Grover's algorithm is very sensitive to noise. Here we present a noise-tolerant method that exponentially improves the noise threshold of Grover's algorithm. We present a lower bound for average fidelity of any quantum circuit with O(log D log D) cost under time-independent noise, where D is the dimension of Hilbert space. According to this bound value, we determine the number of iterates which will be applied in Grover's algorithm. Numerical simulation shows that the noise threshold of quantum advantage of Grover's algorithm by our noise-tolerant method is improved by an exponential factor with qubit amount rise.