Developing a LTE Localization Framework using Real Network Data towards RAN Optimization through Context Knowledge

The exponential growth of the network elements and data traffic exchange in the last few years elevated the need of network providers for optimized and cost-efficient solutions regarding network management and monitorization. Solutions such as drive-tests (DTs) are becoming extremely expensive with the vast extension and complexity of nowadays mobile networks. Therefore, this paper provides a solution for optimized networkcontext knowledge acquisition, towards the self-organizing networks (SONs) concept. The presented framework incorporates an entire scheme for network Traces processing and positioning, based on network measurements and fingerprinting techniques. This framework enables a series of different use cases for network management and optimization, with real-time data processing capabilities within the network Traces collection interval (15 minutes), and achieving a median positioning error of 90 m.