Traffic Pattern Prediction in Cellular Networks

Download Now Free registration required

Executive Summary

Early detection and rapid resolution network congestion can considerably improve network capacity. Consequently, much research has been carried out on predicting traffic patterns in 3G networks. This paper introduces an access point centric approach implemented based on the Order-k Markov model. The feasibility of this Order-k Markov predictor is evaluated, and the prediction performance of predictors with different order values is compared. Finally, the authors conclude that the Order-4 Markov model is competent for both the one-step-ahead and multi-step-ahead traffic pattern predictions.

  • Format: PDF
  • Size: 197 KB