Multiple Radio Access Technologies (Multi-RAT) heterogeneous Networks (HetNets) have been proposed as one of the main solutions to comprise the enormous growth of data traffic in fifth generation (5G) mobile networks. In this paper we consider RAT selection problem in Multi-RAT HetNets where user equipment (UE) has to choose between Long term Evolution (LTE) and Wireless Local Area Network (WLAN) radio access technologies. In this context, the RAT selection is modeled as a UE association process and formulated as an optimization problem aiming to maximize a certain UEs' weighted objective function in terms of user equipment achievable data rates and the security levels provided by different base stations (BSs). The weighting parameter is decided to reflect the orientation of UEs in which higher weighting value for the security level, means that UEs tends to select the BSs that can provide best security levels rather than achieved data rates. A many-to-one matching game-based association approach is then proposed and applied to capture optimal solution for UE association problem. Our simulation results are compared to the well-known WLAN first association algorithm, the results show the outperformance of the proposed matching game based association approach in terms of the achieved system throughput and rate outage probability at different requirements of security levels.