Compensation for multi-path interference using pilot symbols

• US 5,414,734 A
• Filed: 01/06/1993
• Issued: 05/09/1995
• Est. Priority Date: 01/06/1993
• Status: Expired due to Term

• Current Assignee: Quarterhill Inc. (f/k/a Wi-LAN Inc.)
• Original Assignee: Glenayre Electronics Incorporated (EDCI Holdings Incorporated)
• Inventors: Marchetto, Robert F., Stewart, Todd A., Ho, Paul K.-M.
• Primary Examiner(s): Chin, Stephen
• Assistant Examiner(s): Bocure, Tesfaldet
• Application Number:
  US08/001,061
• Time in Patent Office: 853 Days
• Field of Search: 375/99, 375/100, 375/102, 375/40, 375/58, 375/94, 375/103, 375/12, 375/14, 375/11, 455/303, 455/304, 455/306, 455/65, 455/521, 364/724.1, 364/577, 370/110.1, 370/110.2, 370/110.3, 370/110.4
• US Class Current: 375/267
• CPC Class Codes: H04B 7/005 Control of transmission; Equalising

Abstract

A method and apparatus for compensating fading and interference in a radio signal. A plurality of pilot symbols are appended to a plurality of data symbols to form successive frames that are modulated at a transmitter. The transmitted modulated signal is subject to loss of data due to simple fading and multi-path and simulcast interference. The received signals are demodulated by a receiver and processed to provide a data signal comprising the data symbols and a pilot signal comprising the pilot symbols. The data signal is delayed for sufficient time to enable channel impulse response estimates to be made of successive blocks of pilot symbols, preferably using pilot symbol blocks that both precede and follow the data symbols in the frame being processed. The channel impulse response estimates for blocks of pilot symbols are buffered and used by an interpolator to determine an interpolated channel impulse response estimate for each data symbol as a function of both the pilot symbols and of predefined channel characteristics. The interpolated channel impulse response estimates are applied to successive data symbols in the delayed data signal, enabling the data to be decoded, compensating for fading and interference. Interpolation using predefined channel characteristics based on worst case conditions substantially improves the bit error rate (BER) for the data recovered, compared to the prior art.

Claims

1. Circuitry for compensating a receiver for fading of signals propagating as frames of data from at least one transmitter that includes pilot symbol generation circuitry that provides a predefined plurality of pilot symbols in each frame of data transmitted to the receiver, said circuitry comprising:

(a) separating means, coupled to the receiver for input of a received signal, for separating the plurality of pilot symbols from a plurality of data symbols in each frame of data, producing a pilot symbol signal comprising said plurality of pilot symbols and a data signal comprising said plurality of data symbols, said pilot signal and said data signal being separate from each other;

(b) delay means, coupled to receive the data signal, for delaying the data signal from a current frame until after the pilot symbol signal from at least one subsequent frame is received, said delay means thereby producing a delayed data signal;

(c) pilot signal processing means, coupled to receive the pilot symbol signal, for determining an estimated channel impulse response for the plurality of pilot symbols in each frame of received signals;

(d) interpolation filter means, coupled to receive the estimated channel impulse response for each frame of data and including storage means for storing an estimated channel impulse response from at least one prior frame, said interpolation filter means interpolating values of the estimated channel impulse response between the current and prior frames to determine, as a function of channel characteristics, an interpolated channel impulse response for each data symbol in a frame; and

(e) decoder means, coupled to receive the interpolated channel impulse response and the delayed data signal, for recovering the data transmitted to the receiver as a function of the interpolated channel impulse response and the delayed data signal, thereby substantially compensating for fading and interference in the received signals.

2. The circuitry of claim 1, wherein the separating means comprise timing means that extract a portion of each frame for the pilot symbol signal, as a function of a predefined time interval during which the plurality of pilot symbols in each frame are temporally present in the received signal, and then extract a temporal remainder of each frame of the received signal to produce the data signal.

3. The circuitry of claim 1, wherein the separating means, the delay means, the pilot signal processing means, the interpolation filter means, and the decoder means comprise a digital signal processor.

4. The circuitry of claim 1, wherein the fading is due to an interference of at least two signals traveling over different propagation paths to the receiver, and wherein the interpolation filter means comprise matrix operator means for determining the interpolated channel impulse response as a function of a vector product of a vector representing the channel characteristics and a vector of channel impulse response estimates derived from the received pilot symbols.

5. The circuitry of claim 1, wherein the decoder means comprise an equalizer and a decision device.

6. The circuitry of claim 1, wherein said interpolation filter means uses channel characteristics selected from a set including:

(a) a Doppler fading frequency;

(b) relative signal strengths of interfering signals at the receiver;

(c) propagation delay differences between the interfering signals;

(d) frequency offsets between the interfering signals; and

(e) a signal-to-noise ratio of the received signals.

7. The circuitry of claim 6, wherein the channel characteristics are selected based on a worst case scenario for fading and interference between the received signals.

8. The circuitry of claim 1, wherein the decoder means select a data vector for each data symbol from a set of possible values of data vectors, {S}, by determining a minimum value for an expression, D({S}), after evaluating the expression for each data vector in the set, the expression being defined as:

##EQU12## where;

r(i) is the i_th received data symbol;

{W(i)} is the interpolated channel impulse response for the i_th data symbol;

M is a total number of data symbols and pilot symbols in each frame; and

L is a duration for the channel impulse response of the received signals.

9. The circuitry of claim 8, wherein the decoder means comprise a Viterbi decoder, which optimally selects the set {S}.

10. The circuitry of claim 8, wherein the decoder means comprise a Viterbi decoder that determines the set {S} by implementing a reduced complexity form of a Viterbi algorithm, having a suboptimal performance.

11. A method for compensating a receiver for fading of signals propagating as frames of data from at least one transmitter that provides a predefined plurality of pilot symbols in each frame of data transmitted to the receiver, said method comprising the steps of:

(a) separating the plurality of pilot symbols from a plurality of data symbols in each frame of data, producing a pilot symbol signal and a data signal that are separate from each other;

(b) delaying the data signal from a current frame until after the pilot symbol signal from at least one subsequent frame is received, thereby producing a delayed data signal;

(c) determining an estimated channel impulse response from the plurality of pilot symbols in each frame of the received signal;

(d) storing the estimated channel impulse response from at least one prior frame;

(e) interpolating values of the estimated channel impulse response between the current and prior frames to determine, as a function of channel characteristics, an interpolated channel impulse response for each data symbol in a frame; and

(f) recovering the data transmitted to the receiver as a function of the interpolated channel impulse response and the delayed data signal.

12. The method of claim 11, wherein the step of separating the plurality of pilot symbols from each frame of data comprises the step of extracting a portion of each frame for the pilot symbol signal, as a function of a predefined time interval during which the plurality of pilot symbols in each frame are temporally present in the received signal, and then extracting a temporal remainder of each frame of the received signal for the data signal.

13. The method of claim 11, wherein the fading is due to an interference of at least two signals traveling over different propagation paths to the receiver, and wherein the step of interpolating comprises the step of determining the estimated channel impulse response as a function of a vector product of a vector representing the channel characteristics, and a vector of channel impulse response estimates derived from the received pilot symbols.

14. The method of claim 11, wherein the step of interpolating uses predefined channel characteristics selected from a set that includes:

(a) a Doppler fading frequency;

(b) relative signal strengths of interfering signals at the receiver;

(c) propagation delay differences between the interfering signals;

(d) frequency offsets between the interfering signals; and

(e) a signal-to-noise ratio of the received signals.

15. The method of claim 11, wherein values for the channel characteristics are predefined and selected based on a worst case scenario for fading and interference between signals received by the receiver.

16. The method of claim 11, wherein the step of recovering comprises the step of decoding the data transmitted with a bit error rate that over a predefined range, decreases, as a differential delay between interfering signals received by the receiver increases.

17. The method of claim 11, wherein the signals received convey substantially the same data, transmitted substantially simultaneously from different transmitters, said step of interpolating compensating for fading due to interference between said signals.

18. The method of claim 11, wherein the step of interpolating comprises the step of determining, for each data symbol in a frame, a corresponding interpolated channel impulse response, {W}, defined by:

space="preserve" listing-type="equation">{W}={F(opt)}{V}

where;

{F(opt)} is an optimal interpolator matrix determined as a function of the channel characteristics; and

{V} is a vector comprising estimated channel impulse response values derived from a plurality of pilot symbols in a plurality of frames.

19. The method of claim 11, wherein the step of recovering the data comprises the step of selecting a data vector for each data symbol from a set of possible values of data vectors, {S}, by determining a minimum value for an expression, D{S}), after evaluating the expression for each data vector in the set, the expression being defined as:

##EQU13## where;

r(i) is the i_th received data symbol;

{W(i)} is the interpolated channel impulse response for the i_th data symbol;

M is a total number of pilot symbols and data symbols in each frame; and

L+1 is a duration for the channel impulse response of the received signals.

20. The method of claim 19, wherein {W(i)} is equal to a vector product of an optimal interpolator matrix {F(opt)} and an estimated channel impulse response vector, {V}, and {F(opt)} is defined by:

space="preserve" listing-type="equation">{F(opt)}={R(us)}{R(vv)}.sup.-1

where;

{R(uv)} is a correlation matrix;

{R(vv)} is a covariance matrix;

v is a vector comprising estimated channel impulse response values derived from a plurality of pilot symbols in a plurality of frames; and

u is an actual channel impulse response.

21. The method of claim 20, wherein the optimal interpolator matrix {F(opt)} includes L+1 interpolators, each row of said matrix comprising a different interpolator, said step of interpolating comprising the step of multiplying the estimated channel impulse response by each of said L+1 interpolators.

22. The method of claim 20, wherein said plurality of data symbols are grouped together and the same optimal interpolator matrix {F(opt)} operates on each data symbol in a group, thereby reducing storage for interpolator coefficients.