Useful resource parameter for Traffic prediction over GSM

Abstract

The prediction of the voice traffic and their correct version is a assignment for telecom
provider. it's miles important for the fine & provider evaluation of cell network. The take a
look at on this paper subject predicting the voice traffic (Erlang) of cellular network at busy
hour ( five to 7 pm) of an afternoon that is nonlinear & dynamic. This voice visitor also
depends on many different non-linear &dynamic parameters which high-quality expect for
voice site visitors. The information is collected from nice & service record (QOS) of telecom
provider. Correlation evaluation is used to pick proper input variables from QOS document.
Feed forward again propagation algorithm is proposed to make visitors prediction at busy
hours on every day basis.