M Improving Mobility Prediction Using Data Mining

With the increase in the popularity of mobile devices, WLAN infrastructure planning has become significant for maintaining desired quality of service. Movement pattern of the user plays a key role in maintaining quality of service. Mobility prediction involves predicting the mobile device’s next access point as it moves through the wireless network. In this paper, we propose a new method for feature extraction and use Neural network classifier to evaluate the classification accuracy. We evaluate our hypotheses using one month syslog data of Darthmouth college mobility traces available online, to extract mobility features. Our proposed methodology was implemented and achieved a precision of 0.86 and recall of 0.82.