Theoretical Analysis of REM-Based Handover Algorithm for Heterogeneous Networks

Handover has been a widely studied topic since the beginning of the mobile communications era, but with the advent of another generation, it is worth seeing it with fresh eyes. Data traffic is expected to keep growing as new use cases will coexist under the same umbrella, e.g., vehicle-to-vehicle or massive-machine-type communications. Heterogeneous networks will give way to multi-tiered networks, and mobility management will become challenging once again. Under the current approach, based uniquely on measurements, the number of handovers will soar, so will the signaling. We propose a handover algorithm that employs multidimensional radio-cognitive databases, namely radio environment maps, to predict the best network connection according to the user's trajectory. Radio environment maps have been extensively used in spectrum-sharing scenarios, and recently, some advances in other areas have been supported by them, such as coverage deployment or interference management. We also present a geometric model that translates the 3GPP specifications into geometry and introduce a new framework that can give useful insights into our proposed technique's performance. We validate our framework through Monte Carlo simulations, and the results show that a drastic reduction of at least 10% in the ping-pong handovers can be achieved, thus reducing the signaling needed.