Tutorial on data science for wireless networks research

Data science, also referred to as data driven research, is research that puts a strong emphasis on starting from large datasets to solve a specific problem. A simple definition could be that data science enables discovering and extracting new knowledge from data. To the present, data science has gained popularity and was successfully applied in several areas like sales/marketing, medicine, banking/finance, etc.

Due to their unpredictable nature, wireless networks are an interesting application area too for data science because they are influenced by both natural phenomena and man made artifacts. Whereas well-defined aspects of wireless systems such as algorithm behavior, propagation on a specific type of channel, etc., can be modeled in simulations, the functioning of the overall systems is difficult to simulate. Due to these limitations, a number of wireless system behaviors can not be identified and/or explained based on traditional simulations-based approach alone.

In order to perform data driven research in a specific field, as are wireless networks, also domain specific knowledge is needed besides computer science and statistical knowledge. However, due to already established practices of pursuing research in wireless networks (using simulation or model based analysis), some attempts of applying data science to wireless networks research do not follow the best practices as defined in the data science community.

This repository provides a set of scripts and results that were used in the tutorial paper on the correct methodology of using data science in wireless networks research. The paper can be found on this link: http://www.mdpi.com/1424-8220/16/6/790/htm