Scientists from iMinds - University of Antwerp created a unique algorithm that predicts how mobile users will move from one mobile antenna to another. Simulations showed that – based on those predictions – unnecessary switching between antennas can be reduced by up to 20%, thus also reducing the risk of temporary mobile connectivity loss. As part of SEMAFOUR as well, researchers from iMinds - Ghent University contributed to the development of an intelligent system that seeks to automatically connect users to the best performing Wi-Fi or 4G antenna in their neighborhood. In specific scenarios, this technology can increase mobile users’ average data rates up to 85%.
A predictive model for reducing unnecessary handovers
High-speed mobile broadband based on 3G, 4G and (going forward) 5G technology requires the installation of large quantities of antennas, with each antenna covering a limited area (or cell). But people are constantly on the go, and continuously move from one cell to another (often even at high speeds, which is the case on highways for instance).
But when our smartphone has to switch quickly between antennas (the so-called handover), there is a risk that we lose our mobile connection. Moreover, those handovers are pretty expensive for telecom operators as they cause quite some communication overhead. Researchers from iMinds - University of Antwerp now propose a new approach to address that problem.
"Within the scope of the European SEMAFOUR project, to which researchers from Ericsson and Nokia participated as well, we developed a smart algorithm that predicts the movements of a mobile phone user on the basis of several parameters," says Bart Sas (iMinds - University of Antwerp). "By assuming that people follow logical patterns through the streets and roads in a cell, we leverage historical user data to predict how people will move through those cells."
"Based on that information – and our predictive algorithm – we can tell a smartphone with which antenna it should connect to avoid short (burdensome and unnecessary) handovers. Simulations indicate that by means of this approach we may reduce the number of unnecessary handovers by up to 20%," he adds.
In other words: this Flemish technology presents new ways to make telecom networks more (cost)-efficient, while resulting in more stable mobile connections.
"And as far as we know, this is really a unique approach," emphasizes Prof. Steven Latré (iMinds - University of Antwerp). "To begin with, only a few predictive algorithms can do this; and we are the only ones working with standard 4G measurements, which eliminates the need for additional GPS support. That is important, because only a fraction of smartphone users have GPS enabled all of the time."
Technology that seeks to automatically connect smartphones with the best performing Wi-Fi or 4G antenna
The SEMAFOUR project also investigated how telecom operators can increase the capacity of their 4G networks by automatically connecting users to the best performing Wi-Fi or 4G antenna in their neighborhood.
"Today, this decision is taken by each device independently, based on personal settings," says Prof. Ingrid Moerman (iMinds - Ghent University). "But SEMAFOUR has investigated intelligent technology that automatically optimizes the distribution of users across mobile networks. Tests have shown that this approach can increase average mobile data rates up to 85%."
"It was important to determine whether this technology is supported by, or complementary with, current standards. And following our analysis, we believe that it can be implemented in the short term," says Pieter Willemen (iMinds - Ghent University). "In addition, this is a distributed technology: it allows for the easy addition of new (4G or Wi-Fi) antennas. And that's a major advantage for operators."
About the SEMAFOUR project
The goal of the SEMAFOUR project was to design and develop a unified self-management system, which enables network operators to holistically manage and operate their complex heterogeneous mobile networks. The project developed and validated new powerful SON algorithms. In addition, the project created and implemented an integrated SON management system. The project consortium included European mobile network operators, providers of telecom technology and services (such as Ericsson and Nokia) and research organizations (including iMinds researchers from Ghent University and the University of Antwerp).