Smoothing receiver channel estimates using spectral estimation

Details Smoothing receiver channel estimates using spectral estimation Smoothing receiver channel estimates using spectral estimation

1999-01-07

2003-01-14

Chin, Stephen (Department: 2734)

Pulse or digital communications

Spread spectrum

Direct sequence

C375S150000, C375S232000, C375S267000, C375S343000, C375S347000, C375S349000, C375S350000, C370S290000, C370S320000, C370S335000, C370S437000, C370S468000, C455S059000, C455S065000, C455S067700, C455S134000, C455S226200, C455S231000, C455S303000, C708S206000

Reexamination Certificate

active

06507602

ABSTRACT:

FIELD OF THE INVENTION
This invention relates to receivers for decoding signals received via multiple propagation paths having different propagation delays and, more particularly, to use of filters periodically adapted to center frequencies of variations of numerical estimates of amplitude and phases of the delayed propagation paths.
BACKGROUND OF THE INVENTION
Radio receivers are often used for decoding fading signals with the aid of estimates of the instantaneous propagation channel phase and amplitude characteristics. An exemplary application for such a radio receiver is a cellular phone for decoding signals transmitted by base stations using code division multiple access (CDMA) protocols.
A radio receiver, such as for a CDMA system, receives digitally coded and modulated signals from a transmitter. These signals include known, preselect signal patterns at known time intervals. Using known signal patterns and, optionally, data signals obtained after data decoding, the receiver forms successive estimates of the phase and amplitude or complex value of propagation path characteristics between the transmitter and the receiver. These include estimates for multiple paths in the case of multi-path propagation.
It is desirable to smooth the sequence of successive channel estimates to reduce noise and estimation error. A smoothing filter that has a symmetrical filter response is appropriate when the fading spectrum is symmetrical about zero frequency, as expected with a long term average. However, in the short term, on the order of seconds, the fading spectrum may be asymmetrical due to non-uniform distribution of the angle of arrival of multi-path arrays.
The use of channel estimation from a received radio signal using both known symbols embedded in the signal, as well as unknown information symbols that are decoded by the receiver, are well known in the art. Examples of such receivers are shown in U.S. Pat. Nos. 5,331,666; 5,335,250; 5,557,645; and 5,619,533, and also U.S. patent application Ser. No. 08/305,727, filed Sep. 14, 1994, all of which are incorporated by reference herein. Exemplary receivers using channel estimation specific to CDMA systems are shown in U.S. Pat. Nos. 5,151,919 and 5,218,619, also incorporate by reference herein.
Smoothing of channel estimates may be accomplished using a Finite Impulse Response (FIR) filter having a series of complex coefficients. Discussion on smoothing channel estimates using FIR filter, or autoregression, may be found in “Adaptive Equalization For Mobile Radio Channels” (Licentiate Thesis, Lars Lindblom, Uppsala University 1992, ISSN 03468887), which is also incorporated by reference herein. This paper discusses the benefit of adapting a smoothing filter's characteristics to the fading spectrum of the signal. However, in the prior art the fading spectrum of a signal was assumed to be symmetrical. Over the long term, for example several minutes, the fading spectrum may be symmetrical in accordance with Jake's model for fading in the urban, mobile radio propagation environment. The use of Jake's model and modifications thereof to speed computation during simulations of communications system performance may be found in a paper entitled “Jake's Fading Model Revisited” by Dent et al.,
ELECTRONICSLETTERS
, Jun. 24, 1993, Volume 29, No. 13, page 1162 et seq., which paper is incorporated by reference herein.
Jake's model assumes a uniform angular distribution of reflecting objects around a mobile receiver. The relative Doppler shift of reflected signals arising from different angles relative to the direction of movement varies with the cosine of the angle of arrival. With a uniform angular distribution, the Doppler spectrum is then symmetrical and two sided, having as much reflected energy arriving from behind the mobile receiver with a negative Doppler frequency shift as from ahead of the receiver, having a positive Doppler frequency shift. Rays reaching the receiver from behind have clearly not propagated an equal distance from transmitter to receiver as rays reaching the receiver from the front. However, these delay differences were ignored in the prior art. Jake's model assumed that rays with such delay differences could nevertheless be combined to produce a net fading waveform for a path of delay equal to the mean of these rays. More specifically, delays lying within a plus or minus 0.5 of a modulation symbol period of each other were combined to produce a net fading ray with a mean delay. Delays outside that plus or minus 0.5 modulation symbol period were grouped into a different plus or minus 0.5 symbol window to obtain a different net fading waveform with a different mean delay. The different net fading waveforms with their associated modulation-symbol-space delays were then taken to characterize a multi-path channel. Each of the multiple paths is nevertheless assumed to conform to Jake's fading model, i.e., each path is the combination of rays arriving uniformly from all directions.
In a wide band CDMA system (WBCDMA), modulation symbol intervals are much shorter. This allows multiple propagation paths to be resolved with much finer time resolution. Thus, it is no longer valid to use a Jake's model which adds rays that differ in their propagation delay by even a fraction of a microsecond. This addition was valid only in the context of narrow band FDMA or medium bandwidth TDMA systems. In WBCDMA systems, it is necessary to restrict combination of different rays reaching the receiver to rays that have the same propagation delay from the base station to the mobile station, within plus or minus 0.5 of a CDMA chip duration. In a five MHZ wide WBCDMA system, a chip duration is typically 0.25 microseconds so that plus or minus 0.5 chips is plus or minus 0.125 microseconds, or plus or minus 37.5 meters expressed as a propagation path length variation. It may be shown that rays with the same delay to this order of accuracy must have reflected from objects lying on an elliptical contour having the base station and the mobile station as its foci. These objects are not any longer uniformly angularly spaced around the mobile receiver, nor are they spaced at the same distance from either the mobile station or the base station. Moreover, since the base station lies inside the elliptical contour, if, as is usual, it employs directional transmit antennae, objects around the elliptical contour will not be uniformly illuminated. Consequently, the fading spectrum of a ray of given delay within plus or minus 0.5 chip periods are no longer symmetrical about zero frequency. In addition, the offset from zero frequency of the centroid of the fading spectrum is no longer independent of the direction of motion. Consequently, the assumptions of the prior art used in channel estimation and smoothing of channel estimates are overly pessimistic as regards to the bandwidth of the fading.
A claim made in the published art for WBCDMA signals is that the high time resolution enables resolution of individual reflecting objects such that each resolved ray is a single, non-fading ray, i.e., WBCDMA “eliminates fading”. It is recognized that such “non-fading” rays will come and go, but on the relatively longer time scale of log normal shadowing, which is easier to track. However, each ray has a varying Doppler frequency, which means that its phase still varies at up to the Doppler rate, even if its amplitude varies much slower. Thus, there remains the need to track the varying complex value of the propagation channel in order to effect coherent signal decoding, i.e., with knowledge of a phase reference. Moreover, the complete elimination of fading by resolving small reflecting objects is not achieved except using very large bandwidths, beyond the bandwidths of anticipated WBCDMA systems, which therefore find themselves in the intermediate region of propagation paths that still each comprise multiple rays. Fading models and channel estimation means for these WBCDMA have not been addressed in the prior art.
The present invention is

Affiliated with

Also associated with

Rating

Smoothing receiver channel estimates using spectral estimation does not yet have a rating. At this time, there are no reviews or comments for this patent.

If you have personal experience with Smoothing receiver channel estimates using spectral estimation, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Smoothing receiver channel estimates using spectral estimation will most certainly appreciate the feedback.

Rate now

Comments { 0 }

Profile ID: LFUS-PAI-O-3072760