Accurate estimation of available bandwidth of WLAN is essential in cognitive resource management of home/office networks. Several approaches including probing based estimation and bandwidth estimation based on channel busy time have been proposed. However, most of them may affect to the performance of existing sessions, or do not consider on the additional resources that can impact to the estimation. In this paper, we enhance the resource management system to provide cognitive estimation of available bandwidth with indoor positioning. Available bandwidth is estimated considering the location of mobile node and neighbor hidden/exposed terminals. Moreover, the impact of hidden/exposed terminals and their traffic to the performance of target link is analyzed. The proposed system and network performance are analyzed based on experimental implementation in a real testbed.