The Embedded Systems and Signal Processing lab formed to investigate how embedded processing and sensing systems employing advanced signal processing techniques can improve medical care and enhance lives. Injuries, wounds, diseases, and learning disabilities deny people the freedom and opportunities they crave. By partnering with world class medical research teams, we have access to data and expertise which allows us to help return some of these freedoms. We are developing platforms to monitor the progression of disease, provide feedback to aid in rehabilitation, and even identify actions and postures which can lead to injury.

We explore theoretical properties of our problems and platforms. These problems include power optimization techniques, developing compact models to represent our problems, new techniques for classification in resource constrained environments, and signal processing methodologies for reducing data and identifying key signals. Our ultimate objective is to bridge the gap between theory and implementation.

This research requires an inherently multi-disciplinary approach, exploiting ideas from fields as diverse as pattern recognition, signal processing, and embedded system design. In most cases, we build our systems from scratch which involves hardware and software design. We use the systems to collect data. The design techniques mostly are derived from case study on data, and by exploiting specific properties of the signal processing.

Design Techniques and Signal Processing for Light-weight Embedded Systems

• Optimization Techniques for Light-weight Embedded Systems
  • Low Power Collaborative Signal Processing
  • Constant Model of Communication
• Real-time Signal Processing for Body Sensor Networks
  • Automatic Sensor Calibration
  • Collaborative Signal Segmentation Techniques
  • Feature Analysis
  • Distributed Classification
• Modeling
  • Grammar Construction for Physical Movements in Body Sensor Networks
  • Light-weight Physical Models
• Collaborative Signal Processing for Environmental and Body Sensor Networks

Applications

• Physical Activity Monitoring for Assisted Living
• Locomotion Monitoring and GAIT analysis
• Stress Evaluation in Soldiers
• Lower Back Injury Monitoring System for Warehouse Workers
• Sport Training Systems
  • Golf
  • Baseball
• Light-weight ECG Analysis for Angina Pectoris