OMCS-HO: Optimal Mobility Model and Cell Selection Scheme for Handover Management in 5G Small Cells

With the accessibility and popularity of wireless services worldwide, mobile connections and applications are expanding at an unprecedented pace, increasing the demand for traffic data. 5G networks are able to cope with high data traffic requirements by managing handovers (HOs) effectively and efficiently. These solutions, however, result in an astronomical number of handovers, resulting in an increase in unnecessary handovers and drop in probability. As mobile users move from cell to cell within the same registration area, HO is an important Quality of Service (QoS) parameter. Also, dense or very dense proliferation of small cells can cause many problems such as delay, HO failures, frequent turnover and ping pong effect. The objective of this work is to propose an optimal mobility model and cell selection scheme to improve user mobility robustness (OMCS-HO) in 5G-HetNets through seamless HO and cell selection. A modified Blowfish Optimization (MPO) algorithm is used to create an optimal motion model by dividing the local region into multiple motion regions. To reduce the effect of horizontal HO, we design an optimized deep belief neural network (O-DBNN) to predict the future movement of mobile users based on the history of their neighbors. For cell selection, we use a local binary search algorithm (LBS) based on various network characteristics and mobile user movements, which selects the best optimal base station. According to simulation results, our OMCS-HO scheme minimizes the number HOs and link failure probability; maximizes the energy efficiency and throughput.