Author: Mandar N. Kulkarni, Graduate Student, UT Austin

Contact: mandar.kulkarni@utexas.edu

Code/MAT files available at https://drive.google.com/open?id=0B5S48qVlTxGGUzlJUW1UdzhWZHc In order to use these codes or MAT files, please cite the following paper:

M. N. Kulkarni, S. Singh and J. G. Andrews, "Coverage and Rate Trends in Dense Urban mmWave Cellular Networks", IEEE Globecom, Austin, Dec. 2014.

OR

S. Singh, M. N. Kulkarni, A. Ghosh, J. G. Andrews, "Tractable Model for Rate in Self-backhauled mmWave Cellular Networks", IEEE JSAC, Oct. 2015.

This folder contains the following:

1) ManhattanCode.m, ChicagoCode.m: Codes for extracting and processing data from the respective shape files of the Manhattan and the Chicago regions into MAT files that can be used for system capacity analysis of cellular networks. Instructions on how to modify the codes for different regions is given in these scripts. PerimeterPolygon.m is a helper MATLAB function that finds perimeter of a polygon. It should lie in the same folder as ManhattanCode.m and ChicagoCode.m.

2) ManhattanDowntownRawData.mat and ChicagoDowntownRawData.mat: Raw data directly extracted from shape files. Here is the procedure to generate such MAT files for other regions. These can be directly processed further using the above two script files.

a) Download the shape files of building perimeter data from the city websites. For example: Chicago or Manhattan.

b) Now extract into MAT files using the following command: "S = shaperead(filename);"

Warnings:

a) Sometimes the building shape files are too big for your MATLAB to read - may not be able to handle so much memory! In that case use a software like QGIS to save a subpart of the shapefile as another shapefile of smaller size. Then use the above mentioned MATLAB command for extracting into MAT file.

b) It must be remembered that all distances in these shapefiles are by default in foot (generally). This must be verified before using the scripts provided in this folder before using for a new urban region (has been verified for the Manhattan and Chicago links provided herewith).

3) LOS_NLOS.m: Function for deciding whether a link is LOS or NLOS (based on whether the link is intersected by any building). Here seg2poly.m is a helper MATLAB function and should be in the same folder as LOS_NLOS.m in order to run that function.

For further assitance with using these codes please contact the author.