Machine Learning to advance sensing in communication platforms

Partner Call open until: Q2 2021

Project Start: Q4 2020


The goal of the project is to develop hybrid machine learning (ML) and signal processing (SP) algorithms to process active sensing data for classification and regression tasks.

The main objectives and topics of investigation are:

  • Low complexity classification and regression algorithms for radar, ultrasound, time of flight sensors.
  • Time series analysis for prediction systems based on low dimensional sensor data for industrial and consumer applications
  • Hybrid signal processing and ML methods for on demand sensing applications
  • Integration of ML inference on low power sensor platforms with communication capability

Expected results

  • ML models to process sensor data on demand
  • ML models to perform classification, regression, segregation, and detection deployed on Edge devices
  • Hybrid ML and SP algorithms for communication, system identification and modeling, impairment compensation and prediction systems.