Phase Estimation and Control of Coherent Electromagnetic Sources
Phase Estimation and Control of Coherent Electromagnetic Sources
Modern technologies such as laser-based manufacturing, directed energy systems, and high-resolution imaging rely on multiple light sources that emit waves in perfect synchronisation. The way these waves combine determines the resulting beam pattern or intensity distribution at the target. However, even small disturbances or noise can disrupt this delicate phase relationship, leading to loss of focus, unwanted interference patterns, or reduced system performance.
Traditional systems rely on complex optical setups that require expensive, high-bandwidth photodetectors and precise mechanical alignment. These approaches also scale poorly as the number of sources increases, making them costly and cumbersome for large arrays.
To overcome this challenge, this project develops a machine learning based system for estimating and controlling the phase state of an array of coherent electromagnetic sources.
We propose:
A compact sensing architecture employing a minimal, strategically arranged array of photomultipliers to capture interference patterns from coherent sources, significantly reducing sensor requirements compared to conventional dense arrays.
A deep learning–based phase estimation framework that accurately reconstructs the phase of individual sources from sparsely sampled measurements.
Integration of the model into a closed-loop control system that autonomously adjusts source parameters to realize desired interference or beam patterns.
A System and Method for Controlling State of Array of Coherent Sources of Electromagnetic Waves
Inventors: Yashwanth Mohan Kumar, Harishankar Manoharan, Vishnu Rajasekharan Unni, Nithyanandan Kanagaraj.
Patent No 565068: Granted by the Indian Patent Office to the Indian Institute of Technology Hyderabad.