Cellular network bandwidth improvement using subscribers’ classification and Wi-Fi offloading

Adewale Adeyinka Ajao, Ben Obaje Abraham, Etinosa Noma Osaghae, Okesola Olatunji, Edikan Ekong, Abdulkareem Ademola

Abstract


Cellular networks are highly prone to congestion especially at peak traffic periods. This is compounded by the fact that the blocking probability increases. In this study, a machine learning based subscriber classification along with an adaptive Wi-Fi offloading scheme is proposed to improve the throughput and lower the blocking probability of the network. The proposed subscriber classification was implemented using a back propagation based artificial neural network. The result of the subscriber classification was used to develop an adaptive Wi-Fi offloading algorithm based on bandwidth utilization and system throughput. The developed neural network models are shown to be effective, with 94.6% in one experiment, in classifying a user into user classes or levels based on previous data usage. The levenberg–marquardt (LM) algorithm gave the highest accuracy in categorizing the four classes. A relatively large sample size was used for the neural network training cycle and the resulting neural network was then made to use many neurons in its hidden layer. The implementation of the proposed subscriber classification and adaptive Wi-Fi offloading scheme led to a 20% drop in blocking probability and a 50.53% increase in the system throughput.


Keywords


Bandwidth utilization; Data traffic; Subscriber classification; Throughput; Wi-Fi offloading

Full Text:

PDF


DOI: https://doi.org/10.11591/eei.v11i2.2575

Refbacks

  • There are currently no refbacks.


Creative Commons License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

Bulletin of EEI Stats