Real-time quantum feedback prepares and stabilizes photon number states
Feedback loops are at the heart of most classical control procedures. A controller compares the signal measured by a sensor with the target value. It adjusts then an actuator in order to stabilize the signal towards its target. Generalizing this scheme to stabilize a micro-system's quantum state relies on quantum feedback, which must overcome a fundamental difficulty: the measurements by the sensor have a random back-action on the system. An optimal compromise employs weak measurements providing partial information with minimal perturbation. The controller should include the effect of this perturbation in the computation of the actuator's unitary operation bringing the incrementally perturbed state closer to the target. While some aspects of this scenario have been experimentally demonstrated for the control of quantum or classical micro-system variables, continuous feedback loop operations permanently stabilizing quantum systems around a target state have not yet been realized. We have implemented such a real-time stabilizing quantum feedback scheme. It prepares on demand photon number states (Fock states) of a microwave field in a superconducting cavity and subsequently reverses the effects of decoherence-induced field quantum jumps. The sensor is a beam of atoms crossing the cavity which repeatedly performs weak quantum non-demolition measurements of the photon number. The controller is implemented in a real-time computer commanding the injection, between measurements, of adjusted small classical fields in the cavity. The microwave field is a quantum oscillator usable as a quantum memory or as a quantum bus swapping information between atoms. By demonstrating that active control can generate non-classical states of this oscillator and combat their decoherence, this experiment is a significant step towards the implementation of complex quantum information operations.