This repository focuses on the implementation and analysis of an adaptive detection algorithm as presented in the seminal paper An Adaptive Detection Algorithm. The repository is a work in progress, and we are continuously adding more simulations and analyses.
- Problem Addressed:
- Detection of signals in the presence of unknown Gaussian noise using statistical hypothesis testing.
- Signal detection is performed in one data vector, with another independent set of signal-free data vectors available. These vectors share the unknown covariance matrix of the noise in the former vector.
- Solution:
- A likelihood ratio decision rule (LRT) is derived to address this problem.
- The performance of the LRT is evaluated under:
- Noise-only scenarios
- Signal-plus-noise scenarios
- Simulations:
- A comprehensive comparison is made between the Likelihood Ratio Test (LRT) and the Matched Filter.
- Key Insights:
- Performance metrics and trade-offs between the two approaches are analyzed to highlight their strengths and weaknesses.
We are actively expanding this repository with additional simulations, analyses, and implementations. Contributions and feedback are welcome!