The future wireless broadband technologies will offer bandwidth intensive services such as rich voice and high definition multimedia. The aggregate bandwidth requirement at base station of a wireless broadband network is the sum of bandwidth requirements from various services viz., VoIP, browsing, e-mail, streaming multimedia, interactive gaming etc. Due to variation in rates of arrival of service requests during the course of the day, the aggregate bandwidth required to service the traffic fluctuates during the course of the day. The aim of this paper is to demonstrate the use of certain correlation functions of stochastic point processes known as product densities to deal with the non-stationarity of the problem. Product densities are suitably defined so that they can be applied to broadband communication networks for non-homogeneous Poisson arrival rate. The aggregate bandwidth and its fluctuation are estimated using Random Point Process (a special class of stochastic point processes) model. An algorithm to evaluate expected aggregate bandwidth and its fluctuation during the course of the day is developed based on Random Point Process and product densities. Such a bandwidth estimation model can assist wireless broadband service providers to estimate time dependent radio spectrum requirements at the base station. The strength of the product density technique lies in expressing the moments of the aggregate bandwidth as integrals, numerical computation of which can be carried out with any technical computing software such as MATLAB.