Abstract
In this paper, a mathematical model for handover performance indicators in LTE(long time evolution) wireless networks is proposed with the aim of predicting handover parameters and improving handover reliability and efficiency, considering the impact and interdependency of key handover settings (offset, TTT, UE velocity, distance). The best combination of key parameters can be easily achieved by finding the optimal solution of a nonlinear optimization problem, which results in efficient reduction of target handover (HO) failure rate and ping pong rate. The modification of the model is also presented taking the impact of UE location and speed into account. Simulation results and matching error analysis indicate that the proposed fitting functions have an average match quality of over 90% in terms of HO failure rate and ping pong rate, which is good enough to be used for handover prediction.