No abstract

Traffic prediction based on time series analysis methods that are low-cost and low computational complexity can offer more efficient resource management and better QoS. Although exponential smoothing is such a kind of method, there is a lack of application in cellular networks and data traffic research, especially with the robust development of mobile Internet applications nowadays. Therefore, this study provides a comprehensive research on cellular network traffic prediction using exponential smoothing methods. More cases of traffic including voice and data in different time granularities as well as different domains compared with other studies are considered. Besides, more exponential smoothing methods are simultaneously investigated for different cases of traffic. Our multiple case study approach leads to a more convincing result of choosing the best fit model. Data collected from real commercial cellular networks is used for experiments to make the results more practical and persuasively. In our experiment method, the model has the lowest RMSE value is chosen among three types of method. The experiment results show that exponential smoothing methods outperform multiplicative seasonal ARIMA, which is slower and more complex in computation in all cases, so they should be recommended for traffic prediction.

Wireless traffic prediction plays an important role in network planning and management, especially for real-time decision making and short-term prediction. Systems require high accuracy, low cost, and low computational complexity prediction methods. Although exponential smoothing is an effective method, there is a lack of use with cellular networks and research on data traffic. The accuracy and suitability of this method need to be evaluated using several types of traffic. Thus, this study introduces the application of exponential smoothing as a method of adaptive forecasting of cellular network traffic for cases of voice (in Erlang) and data (in megabytes or gigabytes). Simple and Error, Trend, Seasonal (ETS) methods are used for exponential smoothing. By investigating the effect of their smoothing factors in describing cellular network traffic, the accuracy of forecast using each method is evaluated. This research comprises a comprehensive analysis approach using multiple case study comparisons to determine the best fit model. Different exponential smoothing models are evaluated for various traffic types in different time scales. The experiments are implemented on real data from a commercial cellular network, which is divided into a training data part for modeling and test data part for forecasting comparison. This study found that ETS framework is not suitable for hourly voice traffic, but it provides nearly the same results with Holt–Winter’s multiplicative seasonal (HWMS) in both cases of daily voice and data traffic. HWMS is presumably encompassed by ETC framework and shows good results in all cases of traffic. Therefore, HWMS is recommended for cellular network traffic prediction due to its simplicity and high accuracy.

Abstract-Mobile traffic modeling and forecasting are the key techniques in terms of network optimization and management because better network management can be achieved through improving the forecasting accuracy. While mobile traffic has been studied extensively and proved to be effectively modeled with ARIMA models, the volatility effect in mobile traffic series that results in forecasting errors was seldom mentioned. In this study, a multiplicative seasonal ARIMA/GARCH building procedure is proposed to show that volatility effect appearing in mobile traffic series can be processed by GARCH models. Our proposed procedure combines several evaluating parameters such as Akaike Information Criterion (AIC), Schwarz Criterion (SIC), forecast performance evaluation information and residual correlogram to find out the most suitable model, based on which descriptive statistics are used to get the final choice. This work indicates that the mobile traffic series can be better modeled and forecasted by applying GARCH models based on a multiplicative seasonal ARIMA.

No abstract