Ince their origins, mobile communication networks have undergone major changes imposed by the need for networks to adapt to user demand. To do this, networks have had to increase in complexity. In turn, complexity has made networks increasingly difficult to design and maintain. To mitigate the impact of network complexity, the concept of self-organizing networks (SON) emerged. Self-organized networks aim at reducing the complexity in the design and maintenance of mobile communication networks by automating processes. Thus, three major blocks in the automation of networks are identified: self-configuration, self-optimization and self-healing.
This thesis contributes to the state of the art of self-organized networks through the identification and subsequent resolution of a problem in each of the three blocks into which they are divided.
With the advent of 5G networks and the speeds they promise to deliver to users, new use cases have emerged. One of these use cases is known as Fixed Wireless Access. In this type of network, the last mile of fiber is replaced by broadband radio access of mobile technologies. Until now, regarding self-configuration, greenfield design methodologies for wireless networks based on mobile communication technologies are based on the premise that users have mobility characteristics. However, in fixed wireless access networks, the antennas of the users are in fixed locations. Therefore, this thesis proposes a novel methodology for finding the optimal locations were to deploy network equipment as well as the configuration of their radio parameters in Fixed Wireless Access networks.
Regarding self-optimization of networks, current algorithms make use of signal maps of the cells in the network so that the changes that these maps would experience after modifying any network parameter can be estimated. In order to obtain these maps, operators use predictive models calibrated through real network measurements.
Buscar en Dimension
