18225934. LEARNING APPARATUS, OPTICAL SIGNAL STATE ESTIMATION APPARATUS, AND LEARNING METHOD simplified abstract (NEC Corporation)
LEARNING APPARATUS, OPTICAL SIGNAL STATE ESTIMATION APPARATUS, AND LEARNING METHOD
Organization Name
Inventor(s)
LEARNING APPARATUS, OPTICAL SIGNAL STATE ESTIMATION APPARATUS, AND LEARNING METHOD - A simplified explanation of the abstract
This abstract first appeared for US patent application 18225934 titled 'LEARNING APPARATUS, OPTICAL SIGNAL STATE ESTIMATION APPARATUS, AND LEARNING METHOD
Simplified Explanation
The abstract of the patent application describes a learning model that can accurately estimate a signal state even when a constellation on a complex plane is rotated. The learning apparatus includes an acquisition section that acquires a constellation of a known state optical signal transmitted through optical fiber, a generation section that generates a corrected constellation by rotating the constellation on a complex plane, and a learning section that uses the constellation and the corrected constellation to train a learning model.
- The learning model is trained to estimate a signal state with high accuracy even when the constellation is rotated.
- The acquisition section acquires a constellation of a known state optical signal transmitted through optical fiber.
- The generation section generates a corrected constellation by rotating the acquired constellation on a complex plane.
- The learning section uses the acquired constellation and the corrected constellation to train a learning model.
Potential applications of this technology:
- Optical communication systems: The technology can be used to improve the accuracy of signal state estimation in optical communication systems, leading to better data transmission and reception.
- Signal processing: The learning model can be applied to various signal processing tasks where accurate estimation of signal states is crucial, such as image or speech recognition.
Problems solved by this technology:
- Accurate estimation of signal states: The technology addresses the problem of accurately estimating signal states even when the constellation is rotated, which can occur due to various factors in optical communication systems.
Benefits of this technology:
- Improved accuracy: The learning model trained using this technology can provide high accuracy in estimating signal states, leading to improved performance in various applications.
- Robustness to rotation: The technology enables accurate estimation even when the constellation is rotated, ensuring reliable signal state estimation in challenging conditions.
Original Abstract Submitted
A learning model is trained so that a signal state can be estimated with high accuracy even in a case where a constellation on a complex plane is rotated. A learning apparatus () includes: an acquisition section () that acquires a constellation of a known state optical signal transmitted through optical fiber; a generation section () that generates a corrected constellation obtained by rotating the constellation on a complex plane; and a learning section () that uses the constellation and the corrected constellation to train a learning model.
