Technical Reports

Railway Cognitive Radio to Enhance Safety, Security, and Performance of Positive Train Control

  • 22
  • Feb
  • 2013
AUTHOR: Dr. Ashwin Amanna
SUBJECT: Cellular Telephones/Other Electronic Devices, Positive Train Control, Signal and Train Control
KEYWORDS: Positive Train Control (PTC), Software Defined Radio (SDR), Cognitive Radio (CR), wireless communications
ABSTRACT: Robust and interoperable wireless communications are vital to Positive Train Control (PTC). The railway industry has started adopting software-defined radios (SDRs) for packet-data transmission. SDR systems realize previously fixed components as reconfigurable software. This project developed a railway cognitive radio (Rail-CR) which implements Artificial Intelligence (AI) decisionmaking in concert with an SDR to adapt to changing wireless conditions and learn from past experience. Objectives of the project included developing a concept of operations for wireless link adaptation based on use-case scenarios for packet radio systems, designing and implementing a decisionmaking architecture on an SDR, designing strategies for radio environment observations, defining operational objectives and performance metrics, and designing and exercising a test plan to demonstrate performance under varying conditions. The decisionmaking architecture of the Rail-CR begins with observations of the wireless environment and performance metrics. The architecture enables adaptation to new situations and the capability to learn from past decisions. The Rail-CR was tested under a variety of interference conditions designed to simulate real-world experiences. Results show that a radio operating with no-cognition was unable to mitigate interference conditions causing either significantly high errors or a loss of connectivity. The Cognitive Engine (CE) successfully overcame the interference by changing configurable parameters.