Optimizing Radio Access Networks for efficient massive MIMO operation
The research is within the area of massive MIMO where we try to find new ways to assist mobility management, where integrated perception and learning requires certain degree of autonomy. At the same time relevant data needs to be analyzed and used as statistical input in for important configuration decisions in massive MIMO scenarios. There are many challenges ahead before massive MIMO is fully autonomous system in reality.
We aim for a machine learning approach for efficient operation of cellular networks based on massive MIMO. The many antennas in massive MIMO base stations give access to details in the radio channel nd opens up for better prediction of both small scale behaviour such as user correlation as well as large scale behaviour such as mobility patterns. This in turn can lead to new opportunities with respect to scheduling approaches and handover strategies in order to provide low latency reliable user connection in mixed and dynamic environments.