Laser Physics Letters
AI-driven classical quantum hybrid channel noise adaptive control method
Ruixue Yang, Lin Bi, Xiaozhe Ban, Gopal Verma
Abstract
Abstract In co-fiber communication systems transmitting classical and quantum signals, the quantum information carried by a single photon is easily affected by phase noise in the fiber, leading to quantum state distortion. Traditional fixed control methods cannot respond promptly to this constantly changing interference. To address this, this paper proposes a lightweight, physically constrained quantum control framework. This framework uses physics-informed neural networks as its core to achieve adaptive compensation for time-varying noise. The model can be efficiently deployed on a Raspberry Pi, and in single-qubit manipulation tasks, its average fidelity is significantly better than baseline methods without physical constraints, while the performance on the Raspberry Pi is highly consistent with that on traditional platforms. This work provides a real-time, reliable, and low-power control path for co-fiber quantum communication.