Applying Machine Learning to mitigate RF impairments in wireless communications systems
Abstract
In traditional wireless communications systems, multiple complex digital signal processing blocks are designed to combat the imperfect radio frequency front-end components. However, due to Part-to-part, Voltage, and Temperature variation, an algorithm working well on one device may not work well on the others. Also, it is not efficient as each block needs to regularly run calibration and compensation at different time slots; hence, the system performance is significantly reduced. In this project, it is proposed to utilize machine learning to learn and compensate for RF front-end impairments in wireless communications systems. The results of this project can be used to improve various end-to-end wireless communication systems such as 5G/6G, IoT systems, device to device, self-driving car, etc.
Author: Nguyen Viet Tu
Read more about the author’s publications here
