Non-prioritized channel assignment improvement based on call traffic intensity and artificial neural network

Abstract

The non-prioritized (NP) channel assignment model is characterized by a high call dropping probability (CDP) of handover calls and an increasing mobile call traffic volume due to the proliferation of mobile devices. In this study, the one-dimensional Markovian NP model has been improved upon using an artificial neural network (ANN) as a prediction mechanism of CDP using predicted traffic intensity and channel parameters to assign calls of different types to channels. A simulation comparison of the CDP of existing NP channel assignment with the NP with traffic intensity (CDPT) and with the ANN traffic intensity prediction model (CDPANN) was carried out and the study shows that the CDP was reduced drastically when the NP channel assignment with ANN assisted trained model was used putting signal quality into consideration. The CDPT has reduced CDP by 3%, 15%, and 40%, while the CDPANN has reduced CDP by 6%, 20%, and 50% for signal quality factors of 0.2 (poor), 0.5 (good), and 0.8 (very good) respectively. This study has shown that under varying radio frequency signal quality conditions, the ANN assisted channel assignment approach will help minimise the problem of high CDP associated with NP channel assignment and thereby improve ubiquitous mobile communication.