l ¼fcWb=2 and fh ¼fc þWb=2, and similarly for E. In our investigation we consider a range of values of Wb such that it varies from an arbitrarily small value (a pure sinusoid) to the full-band (7.5 GHz wide) channel. Fade analysis: The bandwidth, Wb, has a strong impact on the local variation of the Eb ,a s is observed fromFig. 1 .I t is apparent that the narrowband channel is far more susceptible to multipath interference and small-scale fading than the UWB channel. This is quantified by considering the standard deviation of this channel for the two bandwidths, which is found to be 4.6 and 0.5 dB for the narrowband and ultra-wideband cases, respectively. Furthermore, the cumulative distribution functions (CDF's) of Eb are estimated for various values of Wb ,a s shown inFig. 2. From the Figure, a 1 Hz channel will, on average, suffer fades deeper than 5 dB for 26% of the realisations, 7% for the 100 MHz channel, 0.07% for the 500 MHz channel, and practically 0% for the 7.5 GHz channel. The progressive convergence
[1]
Raymond Knopp,et al.
Subspace Analysis of Indoor UWB Channels
,
2005,
EURASIP J. Adv. Signal Process..
[2]
David J. Edwards,et al.
Ultra-wideband : antennas and propagation for communications, radar and imaging
,
2006
.
[3]
M. Dohler,et al.
Ultra-Wideband: Allen/Ultra-Wideband
,
2006
.
[4]
Theodore S. Rappaport,et al.
Wireless communications - principles and practice
,
1996
.
[5]
M. Dahleh.
Laboratory for Information and Decision Systems
,
2005
.
[6]
Peter B. Papazian,et al.
A comparison of 1920-MHz mobile channel diversity gain using horizontal and vertical arrays
,
2001,
IEEE Trans. Commun..
[7]
Helmut Bölcskei,et al.
Ultrawideband Channel Modeling on the Basis of Information-Theoretic Criteria
,
2005,
IEEE Transactions on Wireless Communications.