## MI-CG: Numerically Computing Achievable Rates of Memoryless Channels

This repository provides a MATLAB function mi_cg.m to numerically compute achievable rates for memoryless channels. The function uses a conditionally-Gaussian (CG) channel model to obtain a lower bound on the achievable rate of the true channel. The method is well-known, and it is explained in [this short document](https://mediatum.ub.tum.de/node?id=1533663). Two example scripts that compute several achievable rate curves are also provided.

### Citation

This software and the accompanying document are meant as a tutorial to get started with mutual information as a numerical figure of merit for a communications channel. The software is provided under the open-source [MIT license](https://opensource.org/licenses/MIT). If you use the software in your academic work, please cite the accompanying [document](https://mediatum.ub.tum.de/node?id=1533663) as follows:

> F. J. Garcia-Gomez, “Numerically computing achievable rates of memoryless channels,” TUM University Library, 2019. [Online]. Available: https://mediatum.ub.tum.de/node?id=1533663 

The corresponding BibTeX entry is
```
@article{garcia2019numerically,
	author = "Francisco Javier Garcia-Gomez",
	title = "Numerically Computing Achievable Rates of Memoryless Channels",
	year = "2019",
	journal="TUM University Library",
	url={https://mediatum.ub.tum.de/node?id=1533663}
}
```

### Acknowledgment

This work was supported by the German Research Foundation (DFG) under Grant KR 3517/8-2.