approach

group1: anmoal

• figure out symbol rate
• matched filtering
• sample

group2: chris, emre

figure out modulation type and order by looking at this scatter plot

considerations:

• can use ML or classical methods (using probability models)
• in the program itself:
  • use probabilities
  • maybe have it so that it can take work with a portion of the data and later efficiently incorporate more data if it is needed

emre

complete group2/synthetic_data.py

investigate which intializiations of k-means works best

• do initialization with points circularly arranged? or just normal constellations; make sure to plot it on the scatter plot before sending it thorugh k-means
• after getting the centroids from k-means, plot it on the scatter plot

k-means

https://docs.scipy.org/doc/scipy/reference/cluster.vq.html

https://en.wikipedia.org/wiki/Determining_the_number_of_clusters_in_a_data_set

google: determine k in k means clustering https://www.kdnuggets.com/2019/08/introduction-image-segmentation-k-means-clustering.html

chris

neural network

convert samples to a density gradient grayscale image

https://stackoverflow.com/questions/20105364/how-can-i-make-a-scatter-plot-colored-by-density-in-matplotlib

https://stackoverflow.com/questions/33282368/plotting-a-2d-heatmap-with-matplotlib

https://stackoverflow.com/questions/36957149/density-map-heatmaps-in-matplotlib

https://towardsdatascience.com/objects-counting-by-estimating-a-density-map-with-convolutional-neural-networks-c01086f3b3ec

resources

• “Rohde Schwarz - YouTube” https://www.youtube.com/channel/UCoJfm2BU72j699FH3IUr3mg
• “Understanding Error Vector Magnitude (EVM) - YouTube” https://www.youtube.com/watch?v=rMVAQsUudSs
• “Interpreting Constellation Diagrams - YouTube” https://www.youtube.com/watch?v=aQd_zBytid8