Abstract
The telecommunications industry in sub-Saharan Africa is experiencing a notable seismic
change. With exponential technological growth, evolving teledensity, and growing customer
base, telecommunications is at a crossroads. However, the consequence of this growth has been
a continuous degradation in the quality of service (QoS) provided by Long Term Evolution
(LTE) service providers, which has resulted in unsatisfactory experiences for subscribers and
customers. In light of the extant literature and research on pricing and artificial intelligence in
the context of network management, this study proposes the application of the Long Short-Term
Memory (LSTM) model of deep learning to predict the eNodeB peak patterns and develop a
dynamic pricing scheme in LTE networks. Each of the interconnected components that make up
our proposed architectural framework is essential to the dynamic pricing and design process.
This ensures seamless integration of predictive modelling, decision-making algorithms, and
communication with users. In addition to the comprehensive analysis, the model is applied to a
real world 4G LTE traffic dataset to evaluate and examine the proposed framework. Finally, the
result of this study will demonstrate the utility of the scheme to enhance the quality of LTE
network services in sub-Saharan Africa.
Keywords: Long Short-Term Memory (LSTM), eNodeB, Long Term Evolution (LTE),
Dynamic Pricing
Publication Date: 2025-02-13