rApps designed for prediction
Zinkworks Prediction Platform
In the era of 5G, the need for intelligent prediction in cellular networks has never been more critical. Efficient and agile network operations are essential for meeting the demands of this advanced technology. Zinkworks Prediction Platform provides a cutting-edge solution for prediction modelling, offering seamless integration of patented machine learning algorithms with intuitive APIs to deliver exceptional performance.
Zinkworks Prediction Platform offers three powerful rAPPs that leverage the platform’s capabilities:
Network Positioning Prediction
Framework for multi-trajectory forecasts
Network Traffic Prediction
Intelligent prediction of cellular traffic
Quality of Service Prediction
Centrally analyse and predict network KPI's
About NPP
Network Positioning Prediction
With the advent of 5G, the ability to precisely determine the geographic location and indoor positions of connected devices has become a game-changer. Zinkworks NPP utilises a comprehensive framework for processing and generating multi-trajectory forecasts based on positioning data from various network users and devices. It employs a continuous training and evaluation pipeline with a Multi-probabilistic Long Short-Term Memory (MP-LSTM). By predicting the movements and positioning patterns of users, NPP leads to a more intelligent and efficient network.
Benefits
Why choose NPP
How NPP Works
Intelligent & Efficient
- Zinkworks NPP is a comprehensive framework for processing and generating multi-trajectory forecasts using positioning data from various network users and devices
- Combines encoding, decoding, and feature engineering techniques to produce smooth and precise time series data
- Uses a continuous training and evaluation pipeline with a multi-probability Long Short-Term Memory (LSTM) deep learning model
About NTP
Network Traffic Prediction
Zinkworks Network Traffic Prediction (NTP) is an advanced network traffic prediction tool that captures spatial dependencies and non-linear temporal dynamics, adapting to changing network conditions. Intelligent prediction of cellular traffic loads is crucial for efficient network operations in the 5G core network. By incorporating traffic prediction into the network control plane, the network becomes more intelligent and proactive.
Benefits
Why choose NTP
How NTP Works
Anticipating Network Loads
- Zinkworks NTP is a centralized ML graph-based model in line with the latest 3GPP and O-RAN standardizations.
- Zinkworks NTP is a holistic network traffic prediction model, forecasts the traffic of a complete cellular network cluster, with the ability to learn the non-linear temporal dynamic of changing network conditions.
- Integrable with non-RT and near-RT RIC, it provides adaptive prediction based on periodic evaluations.
About QoS-P
Quality of Service Prediction
Zinkworks QoS prediction (QoS-P) is an r-App designed for 5G networks that can centrally analyse and predict the network Key Performance Indicators (KPIs). QoS-P can integrate with orchestration apps to revolutionise 5G technology by offering adaptive profiling, streamlining operations, reducing infrastructure, automating network awareness slicing, and ultimately boosting customer satisfaction.
Benefits
Why choose QoS-P
How QoS-P Works
Centralised Predictions
- Centalised ML graph-based model in line with the latest 3GPP and O-RAN standardisation.
- Offers a holistic QoS prediction of an entire cellular network cluster. Helps to dynamically associate various QoS profiles with the connected UEs in the evolving network conditions.
- Integrate with NTP and non-RT to deliver adaptive predictions.