However, we did not discuss the crosscorrelation
of the different multipath faded signals. It is well known that the issue with the Jakes model
is that there can be significant cross-correlation between the different generated signals [131, 132].
And when these correlated multipath signals are used to characterize a RAKE receiver demodulating
many multipaths or a spatial diversity receiver, then the system BER results will be slightly
worse than the truly uncorrelated model and thus bias results in a possibly conservative manner.
3.8.2 Modified Jakes Fading
In this section we will discuss a suggested method to modify the Jakes model to have the desired
cross-correlation properties. These models will prove to be useful in FSF channels and antenna diversity
receivers. We will briefly present a few modified Jakes models. First we will start with the one
proposed by [135].
By adjusting the multipath arrival angles and the use of some special sequences we have
(3.93)
With independent random phases that are uniformly distributed and the following:
; (3.94)
In this model orthogonal functions (Walsh-Hadamard sequences, Ak(n)) weight the sinusoids before
summing to generate the uncorrelated waveforms.
The next model we will review is the Li & Huang [136].
Pages:
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265