Designing and Implementation of Mobility Prediction Algorithms for Cellular Green Networks

Abstract

In this paper, the user mobility and prediction were designed and implemented for green cellular network to save power consumption by controlling the status of basestation (on or off) depending on the availability of users. Random way point and Levy methods were used to model the individual movement of user , while geographical dependence method was used to model the relative movement (depending on the cell). In prediction side, Polynomial method gives accurate predictions and fast computation time. Predicted velocity, direction and location of a user will help the basestation to make a decision for next time slot and take a chance to stay more in a sleep mode to save more energy. Geographical information is added to tell the user about the constraints and correct the direction of motion in different areas.In simulation, the number of overall cells is 126, cells_used (ON-state) is 68, cells did not use (OFF-state) is 58. The saving in using the cells is %53.96.