Recently, emergent technologies associated with the advent of the 5th generation (5G) use-cases have made it imperative for high demands in broadband radio spectrum as compared to the current 4th generation (4G) technologies and other outdated ones. This paper introduces an efficient-spectrum management based on localization of the primary user (PU) position in overlay cognitive radio networks (CRNs) towards 5G. In a typical network consisting of licensed system embedded with primary users (PUs) and unlicensed system embedded with secondary users (SUs), the SU opportunistically exploits the unused frequency bands (spectrum holes) of the licensed network provided there is no harmful interference to the PUs. Since the network is dynamic and the PUs are considered as moving targets with constant acceleration, the spectrum holes become unstable; hence, the SUs are met with limitations of detecting white spaces; finally resulting the interference. To this end, both Kriging Interpolation and Kalman Filter (KF) techniques for tracking and estimating the PUs position are proposed. The performance work in this paper is evaluated using the MATLAB 2018a simulation environment with respect to received signal strength indicator (RSSI) scheme. It is shown that both KF and Kriging Interpolation schemes outperformed RSSI. Moreover, it is revealed that, the KF leads to the best performance compared the other two schemes as per localization error prediction rate and system throughput; which makes it as the preferable technique to tackle mentioned challenges encountered by the SUs in such environment.