Power-Model

Power Wild Experiments (data for Power model)

This dataset consists of data collected at two locations. It covers power modeling results (Figures 15 and 16) presented in Section 4.4 of the paper.

Folder Structure

Filename	Description
MN-Wild	Data and processing scripts for power experiments conducted at Minneapolis, MN
MI-Wild	Data and processing scripts for power experiments conducted at Ann Arbor, MI
dtr_[tm/vz].py	Python script to run decision tree regression on processed data (tm: T-Mobile data collected at Minneapolis, MN; vz: Verizon data collected at Ann Arbor, MI)

Requirements

Here are the software/package requirements. The version number in the bracket indicates the minimum version that our script has been tested on.

• Python 3 (>= 3.7.7)
• Pandas (>= 1.1.3)
• Matplotlib (>= 3.3.1)
• scikit-learn (>= 0.24.0)

Running code

After cloning the repository, navigate to Power-Model folder and run the following command.

python3 dtr_tm.py -d MN-Wild/data-processed/cleaned-logs/ -k t-mobile_nsa -f 1
python3 dtr_tm.py -d MN-Wild/data-processed/cleaned-logs/ -k t-mobile_sa -f 1
python3 dtr_vz.py -d MI-Wild/data-processed/ -k mi-vz-hb -f 1
python3 dtr_vz.py -d MI-Wild/data-processed/ -k mn-vz-hb -f 1
python3 dtr_vz.py -d MI-Wild/data-processed/ -k mn-vz-lb -f 1

For the DTR (decision tree regression) step, we use dtr_vz.py for all the VZ data and use dtr_tm.py for all the TM data. The "f" parameter in dtr.py indicates the feature set (1: TH + SS; 2: TH; 3: SS), the example commands above are using "TH" feature for modeling.