Lewenz, Roger (2005) Temporal variation in the radio environment. Doctoral thesis, Durham University.
Modem digital radio systems using high symbol rates are more susceptible to multipath interference than traditional systems. These systems need to adapt to the propagation environment in order to minimise transmission errors while minimising transmitted power and interference between systems. Mobile terminals will experience variations in the propagation environment as they move. Both fixed and mobile terminals will experience variations in the propagation environment as the environment changes with time. The extent and speed of environmental changes are necessary inputs to the design process for new radio systems. Previous reported measurements aimed at characterising the radio environment have been made over short durations of a few tens of milliseconds. In this thesis measurements over periods of up to 15 seconds are described and the results analysed to quantify longer term variations. The channel sounder developed at UMIST was used for the measurements and the environment was sampled over a number of paths in Manchester and Durham. The data from each 15 second measurement was split into two ensembles, one from the first 7.5 seconds and one from the remaining 7.5 seconds so that statistics from different times could be compared. The number of multipath components, the mean delay and the RMS delay spread were extracted for each 20 msec interval of each measurement. Cumulative Distribution Functions formed from the data from the two ensembles of each measurement showed differences between the ensembles. The results from Kolmogorov-Smirnof tests confirmed the differences for all three statistics. Attempts to model the variation of the statistics by fitting Normal and Weibull distributions to the data failed in most cases. PDFs of the RMS delay spread were multimodal suggesting that insufficient data was available. It has been concluded that there is temporal variation in the radio environment but that more measurement data will be needed for the variation to be characterised.
|Item Type:||Thesis (Doctoral)|
|Award:||Doctor of Philosophy|
|Copyright:||Copyright of this thesis is held by the author|
|Deposited On:||09 Sep 2011 09:53|