Ericsson RET Cell Shaper rApp
Ericsson | Network Optimization
Remote Electrical Tilt (RET) optimization based on AI Reinforcement Learning techniques. Enables continuous tilt optimization at cell level and learns how the performance of each cell reacts to antenna tilt changes. Improves coverage and capacity without degrading the quality of the network. Considers impacts at different layers and technologies when proposing changes.
OPEX reductions: Optimization automation and continuous adaptation to traffic profile
Automates complex cell shaping activities at cluster/market level. Significantly reduces the need of manual tilt activities by automating the whole process of data collection, data processing, tilt proposal and actuation.
Performance: Enhancing user satisfaction and reduce churn
Improving performance in challenging scenarios such as congested areas and high interference to name a few. Analysis is beyond mere coverage hole and overshooting detection as it takes a comprehensive analysis considering impact on neighbors, different layers, and different technologies. Results from a highly congested scenario:
- 11%cluster level increase in downlink throughput
- 7% Cluster level increase in downlink congestion reduction
CAPEX avoidance: Delayed need of capacity expansions
Increases traffic served in congested cells with the same available radio capacity. This is achieved by intelligent traffic shifting from congested cells to neighbors with room to accept additional load without performance impact