TROPICAL
Project leader:
- Panagiotis Papapetrou, Professor, DSV, SU
Researchers
- Amin Azari, Postdoctoral Researcher, DSV, SU
Project period: 2019-01-01 to 2019-09-31
Funding: Huawei (Flagship)
Budget: 100K USD
Description
This project is a pilot study on the applicability of temporal machine learning methods, such as time series prediction and LSTM recurrent neural networks for network traffic identification and estimation. Moreover, it will evaluate the performance of random forests for temporal data series corresponding to network traffic for the task of network traffic identification. Finally, these techniques will then be incorporated to dynamically configure the DRX parameters in a simplified network model, hence providing substantial trade-offs between power reduction and network latency.
The main implementation steps include:
- Time series prediction
- Dynamic configuration of DRX
Implementation
- network traffic dataset - a data sample of our project used for network traffic prediction.
Results
|
User Traffic Prediction for Proactive Resource Management: Learning-Powered
Approaches
In In Global Communications Conference (GLOBECOM), 2019 |
|||
|
Cellular Traffic Prediction and Classification: A Comparative Evaluation
of LSTM and ARIMA
In In International Conference on Discovery Science (DS), 2019 |
People
Panagiotis Papapetrou, Professor
sequential and temporal data mining, explainability, healthcare applications