ML in 5G and beyond networks

Machine learning (ML) as a tool for enabling automation in radio access networks (RANs) is growing more important each year due to the densification of networks and the growth in data consumption.  Driven by improved processing and enhanced software techniques and access to massive amount of data, ML techniques promise to combine simplification with improved performance and efficiency. ML methods allow network operators to solve problems that are challenging with traditional algorithms, optimize several variables jointly, and optimize a sequence of decisions. In order to leverage the potentials of ML in wireless communication systems, a deep understanding of existing systems is vital. Combining ML competence with domain knowledge is crucial for allowing us to solve the right problem in a simplified and efficient way. This presentation will provide an overview of challenges, learnings and opportunities when introducing ML in 5G and beyond networks.