Cognitive radar is envisioned to be the future of remote-sensing systems. The idea proposed by details that a cognitive radar will have the ability to learn about the sensing environment and, over time, optimize returns from targets of interest by altering the transmitted pulse waveform. This paper focuses on the first step of the cognitive process for ionospheric radar observations in the equatorial region. In this region, three plasma irregularities are observed very regularly via radar, namely, Spread F, 150 km echoes, and electrojet. During real-time observations, the first step of the cognitive process entails automatically recognizing the current plasma instability forming in the ionosphere. A preliminary test for the detection and classification of these events is performed off-line using radar observations from the Jicamarca Radio Observatory, located in Lima, Peru. The chosen classification process utilizes a supervised learning algorithm (Expectation-Maximization) to generate a probability distribution function (Gaussian Mixture Model) for each plasma instability based on certain measurable criterion. This paper outlines the training and classification process with the resulting cross validation performance subsequently reported. Results demonstrate the effectiveness of the chosen classifier because of the high cross validation and minimal misclassifications of the events.
All Science Journal Classification (ASJC) codes
- Condensed Matter Physics
- Earth and Planetary Sciences(all)
- Electrical and Electronic Engineering