Wideband Autonomous Cognitive Radios (WACRs) for Spectrally-efficient and Agile Multiband/Multimode Communications

Abstract

The objective of the overall project is to develop a wideband autonomous cognitive radio (WACR) prototype to achieve network-, user- and spectrum-aware communications to address limitations in current T&E practices at ARMYÕs test ranges as well as the increasingly significant spectrum encroachment problem caused by the private sector. The research tasks pursued during this project can be summarized as: 1. Develop a wideband spectrum knowledge acquisition framework for L and S band spectrum so that a cognitive radio can be fully aware of its RF environment. The challenge is to achieve this with realistic software-defined radios (SDRs) that have limited instantaneous bandwidths. The work to be pursued aims to develop efficient spectrum scanning approaches to sense the wide spectrum range of interest within such constraints. 2. Develop an efficient real-time signal detection and classification framework to detect and extract all active signals in L and S band spectrum. Depending on the type of application and the particular spectrum range, the detection of signals may require different approaches. Moreover, the ability to identify the detected signals may also be needed in many situations. In this task, both detection and signal classification approaches that are suitable for real-time implementation are investigated. 3. Develop machine learning and deep learning signal classifiers to separate between radar and communications signals and/or legitimate signals vs. spectrum encroachers in the spectrum bands of interest to ARMYÕs WSMR operations. 4. Develop and implement efficient signal feature extraction mechanisms including cognitive direction-of-arrival (DOA) estimation techniques. 5. Develop cognitive dynamic spectrum sharing protocols for the radios to coexist with other signals in the same spectrum range, through spectrum agility. Depending on the type of RF environment, explore either machine or deep learning approaches to avoid interference through such spectrum agility. 6. Implement the developed spectrum knowledge acquisition framework in a real-time test system using Universal Software Radio Peripherals (USRPs) as the software-defined radio platforms. The goal is to demonstrate cognitive dynamic spectrum sharing operation through real-time, or near real-time, spectrum awareness over 1GHz Ð 6GHz.

Document Details

Document Type
DoD Grant Award
Publication Date
Feb 25, 2019
Source ID
W911NF1710035

Entities

People

  • Sudharman Jayaweera

Organizations

  • Army Contracting Command
  • United States Army
  • University of New Mexico

Tags

Fields of Study

  • Computer science
  • Engineering

Readers

  • Distributed Systems and Data Platform Development
  • Radar Systems Engineering.
  • Radio communications and signal processing.

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks