Predicting Location-Dependent QoS for Wireless Networks
In wireless networks the quality of service (QoS) delivered to an end user will be a complex function of location and time. “QoS Seeker” is a new system which informs a user what location in the wireless network he should move to in order deliver the required QoS for his real-time applications. At the heart of QoS Seeker is a construct known as a “QoS Map” – which is the value of a specific QoS metric as a function of space.If any significant temporal trends are present in the wireless network, then the current QoS Map will be statistically different from a future QoS Map. In this report we investigate the use of adaptive linear filters as a means to predict future QoS Maps from historical QoS Maps. By using the received signal strength (RSS) as the QoS metric, we show that local adaptive filters can deliver very significant performance gains relative to last-measure and moving-average predictors. We also show how global adaptive filters can produce performance gains, albeit at a lower level. These results show that adaptive prediction techniques have a significant role to play in the QoS Map construction.
Unable to display preview. Download preview PDF.
- 1.Malaney, R., Exposito, E., Wei, X., Nghia, D.: Seeking VoIP QoS in Physical Space. NICTA-RAM-01-2004 (2004)Google Scholar
- 3.Stocker, K.E., Gschwendtner, B.E., Landstorfer, F.M.: Neural network approach to prediction of terrestrial wave propagation for mobile radio. In: Proc. Inst. Elect. Eng. -H, August 1993, vol. 140, pp. 315–320 (1993)Google Scholar
- 4.Qualcom-SnapTrack, http://www.snaptrack.com
- 5.Wolfle, G., Landstorfer, F.M.: Adaptive propagation modeling based on neural network techniques. In: Proc. IEEE 46th Vehicular Technology Conf. (VTC), Atlanta, GA, pp. 623–626 (1996)Google Scholar