Invention:
This technology is a general recipe for seeking quantum optimum receivers. The recipe utilizes machine-learning techniques to automatically generate the quantum optimum receiver architecture beyond reach of human design.
Background:
Communication systems include sources, channels and receivers. The receiver distinguishes the difference between each encoded signal and decodes then into original loaded information with an acceptable error rate. Quantum information theory has proved that the lowest physically reachable error rate of decoding is far below receiver’s error rate. This has created the need for an improved receiver. There has been research and development aimed to design a much-improved receiver. The introduction of deep learning as a computational model can address this complex task.
Applications:
- Deep-space communications
- Biological sensing
- Imaging
- Quantum information processing
Advantages:
- Generate architectures beyond human design
- Reduce time spent on designing receiver architectures
Status: issued U.S. patent #11,258,519
