Exploration of daily Internet data traffic generated in a smart university campus

In this data article, a robust data exploration is performed on daily Internet data traffic generated in a smart university campus for a period of twelve consecutive (12) months (January–December, 2017). For each day of the one-year study period, Internet data download traffic and Internet data upload traffic at Covenant University, Nigeria were monitored and properly logged using required application software namely: FreeRADIUS; Radius Manager Web application; and Mikrotik Hotspot Manager. A comprehensive dataset with detailed information is provided as supplementary material to this data article for easy research utility and validation. For each month, descriptive statistics of daily Internet data download traffic and daily Internet data upload traffic are presented in tables. Boxplot representations and time series plots are provided to show the trends of data download and upload traffic volume within the smart campus throughout the 12-month period. Frequency distributions of the dataset are illustrated using histograms. In addition, correlation and regression analyses are performed and the results are presented using a scatter plot. Probability Density Functions (PDFs) and Cumulative Distribution Functions (CDFs) of the dataset are also computed. Furthermore, Analysis of Variance (ANOVA) and multiple post-hoc tests are conducted to understand the statistical difference(s) in the Internet traffic volume, if any, across the 12-month period. The robust data exploration provided in this data article will help Internet Service Providers (ISPs) and network administrators in smart campuses to develop empirical model for optimal Quality of Service (QoS), Internet traffic forecasting, and budgeting.


a b s t r a c t
In this data article, a robust data exploration is performed on daily Internet data traffic generated in a smart university campus for a period of twelve consecutive (12) months (January-December, 2017). For each day of the one-year study period, Internet data download traffic and Internet data upload traffic at Covenant University, Nigeria were monitored and properly logged using required application software namely: FreeRADIUS; Radius Manager Web application; and Mikrotik Hotspot Manager. A comprehensive dataset with detailed information is provided as supplementary material to this data article for easy research utility and validation. For each month, descriptive statistics of daily Internet data download traffic and daily Internet data upload traffic are presented in tables. Boxplot representations and time series plots are provided to show the trends of data download and upload traffic volume within the smart campus throughout the 12-month period. Frequency distributions of the dataset are illustrated using histograms. In addition, correlation and regression analyses are performed and the results are presented using a scatter plot. Probability Density Functions (PDFs) and Cumulative Distribution Functions (CDFs) of the dataset are also computed. Furthermore, Analysis of Variance (ANOVA) and multiple post-hoc tests are conducted to understand the statistical difference(s) in the Internet traffic volume, if any, across the 12-month period. The robust data exploration provided in this data article will help Internet Service Providers (ISPs) and network administrators in smart campuses to develop empirical

Value of the data
The data provided in this data article can be used to accurately predict Internet data traffic in a smart campus environment. Predictions of Internet data traffic will help network engineers to improve the Quality of Service (QoS) of computer networks and also ensure efficient utilization of the networks in a smart university campus [1,2].
Availability of dataset on Internet data traffic obtained from real scenarios will facilitate more empirical research in the areas of computer networking and Internet traffic engineering [3,4].
This dataset is made available to give correct facts and figures on Internet data traffic in a Nigerian university campus that is driven by Information and Communication Technologies (ICTs) [5,6].
Free access to daily Internet data traffic of a period of one year will facilitate the development of empirical prediction models that can be used by Internet Service Providers (ISPs) and Internet subscribers in a smart university campus for effective network planning and traffic forecasting [7][8][9][10][11][12].
Robust data exploration that is performed in this data article will help the university network administrators to gain useful insights about the traffic peak and off-peak periods. Also, the descriptive statistics, frequency and probability distribution plots, correlation analysis, ANOVA test and multiple post-hoc test results will give better understanding of the relationships between the Internet data download traffic and the Internet data upload traffic in a smart campus [13][14][15].

Data
Ubiquitous access to reliable Internet services is pivotal to achieving sustainable smart education in university campuses [16][17][18]. Accurate Internet data traffic prediction models are required for computer network planning and forecasting to guarantee efficient Quality of Service (QoS) in enterprise computer networks and applications. However, computer network planning are usually carried out based on theoretical formulations and simulations due to paucity of empirical data from real life scenarios. In this data article, a robust data exploration is performed on daily Internet data traffic in a smart university campus for a period of twelve consecutive (12) months (January-December, 2017).
For each month, descriptive statistics of daily Internet data download traffic and daily Internet data upload traffic are presented in tables. The mean, median, mode, standard deviation, variance, kurtosis, Skewness, range, minimum, maximum, and sum of the daily Internet data traffic download and upload for January-December, 2017 are presented in Tables 1 and 2 respectively.

Experimental design, materials and methods
A robust data exploration was performed on daily Internet data traffic in a smart university campus for a period of twelve consecutive (12) months (January-December, 2017). For each day of the one-year study period, Internet data download traffic and Internet data upload traffic at Covenant University, Nigeria were monitored and properly logged using an open source software, FreeRADIUS, Radius Manager web application, and Mikrotik Hotspot Manager. FreeRADIUS software was installed in Linux Operating System (OS) for authentication, authorization, and accounting services. Radius Manager Web application was used to add users, to edit and create cards, and to harvest data in a more user-friendly format. Mikrotik Hotspot Manager was used to integrate the smart campus network to the enterprise edge. Statistical computations were done using the Machine Learning and Statistics toolbox in MATLAB 2016a software. Boxplot representations of the daily download traffic and the daily upload traffic for the 12-month period are shown in Figs. 1 and 2 respectively.

Data exploration
Time series plots are provided to show the trends of data download and upload volume within the smart campus throughout the 12-month period.      Table 3. Estimates and standard errors of download data traffic distribution parameters for the six models are presented in Table 4. Similarly, the distribution fitting parameters for upload data traffic (January-December, 2017) based on the six distribution models are presented in Table 5. Estimates and standard errors of download data traffic distribution parameters for the six models are presented in Table 6.
Furthermore, Analysis of Variance (ANOVA) and multiple post-hoc tests are conducted to understand the statistical difference(s) in the Internet traffic volume, if any, across the 12-month period. The results of the ANOVA test and the multiple post-hoc test conducted on download data traffic are presented in Tables 7 and 8 respectively. Likewise, the results of the ANOVA test and the multiple post-hoc test conducted on upload data traffic are presented in Tables 9 and 10