SliceNet are delighted to announce our Exhibition & Demonstration booth at EuCNC2019 which will highlight important aspects of network slicing for Smart Grid, Smart City and Smart eHealth use-cases. SliceNet will demonstrate network slicing across multiple administrative domains, meeting the requirements from the management and control planes. This network slicing facilitates early and smooth adoption of 5G slices for verticals to achieve their demanding use cases and manages the QoE for slice services. The SliceNet demonstrations will showcase the Machine Learning (ML)-enabled cognitive network slicing management capabilities in the following vertical use cases:
Smart Grid (Predictive fault management) demonstration explores a dataset provided by the biggest Portuguese Service Provider for the Machine Learning (ML) training. The demo features ML model creation through TensorFlow framework and using a Dense Neural Network (DNN) algorithm, ML model deployment, and ML model execution for RAN slice failure prediction in an execution environment.
Smart City (noisy neighbour) demonstration integrates R&D efforts from two 5G PPP projects SliceNet and MATILDA to show that inter-slice conflict in resources is mitigated. It onboards a 5G Smart City application in the Marketplace, deploys two slices and provides an ML model that monitors slice-level metrics and applies cognitive method to predict service-level degradation and pinpoint its origin.
Smart eHealth demonstration shows eMBB slice composition and onboarding and how MEC with hardware acceleration can assist in continuous collection, processing and streaming of patient data to speed up diagnosing stroke through an ML-based MEC App. Furthermore, it presents mobility management in slice-based networks, highlighting seamless on-demand handovers, monitoring and configuring mobility parameters for SDN-based handovers, and a novel plug & play execution environment for network slices customisation.
SliceNet partners Eurecom, Universitat Politecnica de Catalunya, Orange, DellEMC and NextWorks will support the demos at the booth.