Introduction

This project is inspired by the annual reports issued by the “National Instance of Telecommunications of Tunisia”. I used data from 2017 annual report for 2G and 3G networks ( Link). I used these main Key Performance Indicators (KPIs):

• For Voice Services:
  • Blocking rate (BR,%): Ratio of the number of blocked call attempts to the total number of call attempts made. A call attempt is considered blocked if the ring back is not received after 20 seconds. The call attempts are spaced by a variable time interval equal to at least 30 seconds.
  • Interruption rate (IR, %): Ratio of the number of 2-minute calls cut off before the end of this time to the total number of unblocked calls. - Mean Opinion Score (MOS): The score assigned to a call following an evaluation of its auditory quality using a scoring algorithm based on calls with an average duration of 2 minutes.

• For Data Service:

  • Access success rate (ASR, %): The ratio of the number of access attempts to a set of websites or a set of HTTP streaming videos that were not blocked to the total number of access attempts.

  • Average 2G/3G browsing speed (ABS , kbps): The ratio of the sum of bits received during an HTTP browsing session to the duration of that session.

  • Successful download/send rate (SD/SR,%): The ratio of the number of successful downloads/sends to the total number of downloads/sends.

  • Average file download/sending rate (AFD/SR, Mbps): Ratio of the sum of bits received/sent during a File Transfer Protocol (FTP) session to the duration of that session.

Data will be plotted by gouvernorate (24) and by Internet Provider Service (IPS). In Tunisia, we have 3 main IPS for cellular data: - “Tunisie Telecom” - Ooredoo - Orange Tunisie

Data preprocessing

• In order to represent the data on a leaflet map I used “geopy” (Python package) in order to extract longitude and latitude.

import pandas as pd 
from geopy.geocoders import Nominatim

#Importer les données
df = pd.read_excel("2g3g-f.xlsx")

#Définir les régions et les opérateurs.
op = []
reg = []
for i in range(len(df)):
    s = df['gouvernorat'][i]
    l= s.split(": ")
    if(l[1] == "TT"):
        op.append("Tunisie Télécom")
    if(l[1] == "OT"):
        op.append("Orange Tunisie")
    if(l[1] == "OO"):
        op.append("Ooredoo")
    reg.append(l[0])
    
df.gouvernorat = reg
df['opérateur'] = op

# Chercher les longitudes et latitudes.
geolocator = Nominatim(user_agent="https://maps.google.com/")
lat = []
long = []
for i in range(len(df)):
    try:
        location = geolocator.geocode(df.gouvernorat[i])
        lat.append(location.latitude)
        long.append(location.longitude)
    except:
        lat.append("NAN")
        long.append("NAN")
        
df['Longitude'] = long
df['Latitude'] = lat
pd.DataFrame.to_csv(df,"data.csv")

Shiny application

• In the application you, you’ll find 3 tabs:
  • The first one is the Tunisian map that contains the presented informations for every governorate and for every IPS.
  • The second tab contains data preview.
  • Finally, you’ll find the explanations of the KPIs.

• Link to the Shiny App.

• Link to the Github Repo.