Artificial Intelligence Based Feature Selection and Prediction of Downlink IP Throughput | IEEE Conference Publication | IEEE Xplore

Artificial Intelligence Based Feature Selection and Prediction of Downlink IP Throughput


Abstract:

With the rapid development of artificial intelligence, random forest is widely used in feature selection in wireless communication systems. Due to its ability to efficien...Show More

Abstract:

With the rapid development of artificial intelligence, random forest is widely used in feature selection in wireless communication systems. Due to its ability to efficiently handle large data sets with high dimensions, random forest has become a popular machine learning method. The Downlink IP Throughput is a key indicator reflecting the real business experience of users in wireless communication systems. It is defined as the ratio of the amount of data sent downlink at the RLC layer to the data transmission time. This paper proposes a feature selection method based on random forest to predict Downlink IP Throughput. Before the random forest is used, the feature dimension is reduced in advance by calculating the variance of the features and filtering the features with low variance. Then, the random forest model is used to select the features that are highly sensitive to the Downlink IP Throughput and can mostly affect and reflect the change of the Downlink IP Throughput, and the feature set that affects the Downlink IP Throughput is constructed, and the MLP model is built to predict the Downlink IP Throughput. The model uses real data such as user scheduling information and user throughput measurement information to train and test the prediction model. The experimental results show that the model successfully selects the key features that affect the Downlink IP Throughput, and shows good performance in the Downlink IP Throughput prediction.
Date of Conference: 02-04 November 2023
Date Added to IEEE Xplore: 02 February 2024
ISBN Information:
Conference Location: Hangzhou, China

Funding Agency:


References

References is not available for this document.