Learning-based blockage prediction for robust links in dynamic millimeter wave networks

To employ millimeter wave technology in 5G networks, two inherent challenges need to be addressed in dynamic outdoor environments. Firstly, different types of obstacles can easily block the links. Secondly, the link quality can drop significantly in a mobile environment. It is critical to discriminate between the two different situations to take appropriate actions. Existing work makes the distinction based on RSSI variation measured in a time window, which is very time-consuming, leading to a large volume of data loss to achieve high accuracy. This paper proposes a learning-based prediction framework to classify link blockage and link movement efficiently and quickly. A classifier is trained with data blockage instances using different learning methods. It is used to make a prediction based on diffraction values on different multipath components formed around a receiver. Simulation results for both blockage and link movement show that the prediction framework can predict blockage with close to 90% accuracy. The accuracy of detecting the blockage (not the link movement) is measured in the experiments as well to analyze the feasibility of the method. The prediction framework can eliminate the need for having time-consuming methods to discriminate between link movement and link blockage. The simulations show that our framework does not need a large amount of training data to achieve the desired prediction accuracy. The experiments using commodity millimeter wave radios demonstrate a very high prediction accuracy.