Abstract: The paradigm shift from fixed homogeneous network to mobile heterogeneous network for mobile multimedia applications with QoS (Quality of Service) is growing enormously. The main QoS parameters for mobile multimedia application are high bandwidth and low delay. To achieve satisfactory service delivery for mobile multimedia applications in heterogeneous network, the above mentioned QoS parameter is to be met. Provision of such services in heterogeneous environment invites many research issues. One of the main challenge is the selection of optimal network with QoS since current QoS schemes have two drawbacks such as unawareness of the heterogeneous wireless environment and inefficient utilization of the reserved bandwidth. To solve these problems, we present a novel QoS parameter namely Network Quality Factor (NQF) to estimate the network status. NQF is evaluated by estimating the Call blocking probability (Cbp) of heterogeneous network from its ‘rch’ (recent call history). The estimation of ‘rch’ parameter and Cbp avoids unnecessary hand-off and enhances the effective bandwidth utilization of a cellular network under heavy load condition. The predicted ‘rch’ parameter of a network is then mapped in the proposed QoS mapper sub-module present in QoS broker module to select an optimum network. The network with higher NQF is always termed as optimum network as it has low Cbp and it is always selected. A Bayesian Network Model is assumed for Cbp estimation and simulated using Monte-Carlo Method over various heterogeneous environment conditions. The estimated NQF is then mapped with User Satisfaction Factor (USF) to select the most suitable access network for each session. The modeling and simulation results demonstrate that the NQF based optimum network selection can satisfy user’s QoS requirement and obtain a more efficient use of the scarce wireless bandwidth during heavy traffic load condition.