Cellular Traffic Prediction Using Online Prediction Algorithms

TL;DR

This work tackles real-time cellular network traffic prediction by evaluating online predictors that forego retraining. It contrasts Rolling and Fast LiveStream Prediction (FLSP) across statistical (ARIMA, SARIMA) and deep-learning (LSTM, CNN-LSTM, ConvLSTM) models under synchronous and asynchronous data gathering, with a focus on accuracy, latency, and bandwidth. Key findings show that FLSP reduces reporting bandwidth in asynchronous setups by about half while often improving predictive accuracy and reducing computation, particularly for stateful models like ConvLSTM; synchronous results also favor FLSP, delivering strong accuracy with lower data transfer. The results offer practical guidance for live network optimization and proactive resource allocation in dynamic 5G/6G environments, highlighting the trade-offs between model complexity, memory, and data-gathering strategies.

Abstract

The advent of 5G technology promises a paradigm shift in the realm of telecommunications, offering unprecedented speeds and connectivity. However, the efficient management of traffic in 5G networks remains a critical challenge. It is due to the dynamic and heterogeneous nature of network traffic, varying user behaviors, extended network size, and diverse applications, all of which demand highly accurate and adaptable prediction models to optimize network resource allocation and management. This paper investigates the efficacy of live prediction algorithms for forecasting cellular network traffic in real-time scenarios. We apply two live prediction algorithms on machine learning models, one of which is recently proposed Fast LiveStream Prediction (FLSP) algorithm. We examine the performance of these algorithms under two distinct data gathering methodologies: synchronous, where all network cells report statistics simultaneously, and asynchronous, where reporting occurs across consecutive time slots. Our study delves into the impact of these gathering scenarios on the predictive performance of traffic models. Our study reveals that the FLSP algorithm can halve the required bandwidth for asynchronous data reporting compared to conventional online prediction algorithms, while simultaneously enhancing prediction accuracy and reducing processing load. Additionally, we conduct a thorough analysis of algorithmic complexity and memory requirements across various machine learning models. Through empirical evaluation, we provide insights into the trade-offs inherent in different prediction strategies, offering valuable guidance for network optimization and resource allocation in dynamic environments.

Figures (7)

• Figure 1: Flowchart of FLSP algorithm.
• Figure 2: ML Models with memory: (a) LSTM-DenseNet model, (b) CNN-LSTM model with CNN layers connected in DenseNet architecture, (c) ConvLSTM-CNN model with DenseNet connection between the layers.
• Figure 3: STL decomposition of call volume assuming weekly seasonality.
• Figure 4: Spatial correlation of cells surrounding cell 4445. It shows high correlation of at least 0.86 among the neighbouring cells.
• Figure 5: Online predictions using statistical models and rolling algorithm. Fresh data is fed to the models as a single batch of data with a size of $15$ samples, and predictions are made for two batches ahead (with a size of $2*15=30$ time slots) at each step. (a) Comparison of the predicted and ground truth patterns when using ARIMA model, resulting in an MSE of $513.53$. (b) Predicting the traffic pattern using SARIMA model achieves an MSE of $61.78$.