LEIMER DONALD (US)
| US20050105647A1 | ||||
| US6778130B1 | ||||
| US20030095529A1 |
Claims
I claim:
1. A method for estimating a Doppler spread and a signal-to-noise ratio of a received signal, comprising:
(a) Calculating one or more functions of a carrier component of the received signal at a first time point relative to a pilot signal embedded in the received signal;
(b) Calculating the one or more functions of the carrier component of the received signal at a second time point relative to the pilot signal;
(c) repeating steps (a) and (b) over multiple time periods, each time period being longer than the second time, and accumulating the one or more functions for the first time points and the second time points; and
(d) Deriving the Doppler spread and the signal-to-noise ration based on the accumulated one or more functions calculated at the first and second time points.
2. A method as in Claim 1, wherein the first time point is substantially closer to the pilot signal than the second time point.
3. A method as in Claim 2, wherein the first time point is one symbol interval from the pilot signal.
4. The method of Claim 1 , wherein a product of the Doppler spread and the time difference between the first time point and the second time point is a positive value.
5. The method of Claim 4, wherein the positive value is approximately 0.25.
6. The method of Claim 4, wherein the positive value is less than or equal to 1.0.
7. The method of Claim 1, wherein the one or more functions comprise an even- symmetric linearity function of a sum of the carrier component and the pilot signal.
8. The method of Claim 1, wherein the one or more functions comprise an even- symmetric linearity of a difference between the carrier component and the pilot signal.
9. The method of Claim 1, wherein the even-symmetric linearity comprises an absolute value function.
10. The method of Claim 1, wherein the even-symmetric linearity comprises a squaring function.
11. The method of Claim 1 , wherein the time domain representation of the received signal comprises a sum of the transmitted signal and noise.
12. The method of Claim 1, wherein the received signal is modeled as a Rayleigh faded continuous wave signal.
13. The method of Claim 1, wherein the one or more functions being provided by Jake's model. |
METHOD OF ESTIMATING DOPPLER SPREAD AND SIGNAL-TO-NOISE RATIO OF A RECEIVED SIGNAL
Donald Leimer
CROSS-REFERENCE TO RELATED APPLICATIONS The present application is related to and claims benefit of priority of U.S. provisional application, entitled "Method of Estimating Doppler Spread and Signal-to-Noise Ratio of a Received Signal," serial no. 60/756,013, filed on January 4, 2006; and U.S. nonprovisional application, entitled "Method of Estimating Doppler Spread and Signal-To-Noise Ratio of a Received Signal," serial no. 11/619,178, filed on January 2, 2007. These applications are hereby incorporated by reference in their entirety. For the US designation, the present application is a continuation of the aforementioned U.S. patent application no. 11/619,178.
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to signal processing techniques, hi particular, the present invention relates to estimating Doppler spread and signal-to-noise ratio (SNR) of a received signal.
2. Discussion of the Related Art
hi many communication systems, the received signal often includes a training sequence or a pilot subcarrier. For example, DVB-T/H OFDM signals are examples of such signals. In such systems, it is desirable to estimate both the Doppler spread and SNR of the received signal. The estimated Doppler spread and SNR are used to select optimum values of reception and demodulation parameters that accommodate channel conditions change.
hi an article by David R. Pauluzzi, "A Comparison of SNR Estimation Techniques for the AWGN Channel", IEEE Transactions on Communications; Vol. 48, No. 10, pp. 1681- 1691 (October 2000), numerous methods for estimating the SNR of single-carrier signals in additive white Gaussian noise (AWGN) channels are discussed. These methods can be extended to estimate the SNR of an OFDM signal by treating the OFDM signal as a collection of single-channel signals. Likewise, some of these methods may be extended to estimate the SNR in Rayleigh fading channels, using an appropriately small time interval to estimate the signal statistics. The split-symbol moment estimator is particularly suited for this approach. See, e.g., S.K. Simon and A. Mileant, "SNR Estimation for the Baseband Assembly"; Telecommunications and Data Acquisition Prog. Rep. 42-85 Jet Propulsion Lab.,
Pasadena, California (May 15, 1986); and B. Shah and S.Hinedi, "The Split Symbol Moments SNR Estimator in Narrowband Channels"; IEEE Transactions on Aerospace Electronic Systems; Vol. AES-26, pp. 737-747 (Sept. 1990). U.S. Patent 6,922,452 to Sandberg describes a method for estimating Doppler spread on a single-channel signal by computing the autocorrelation of the received signal and relating its first zero-crossing to the Doppler spread bandwidth. While this method could be applied to individual subcarriers of an OFDM signal, the computation and storage requirements are excessive for a mobile receiver that must consume a minimum of power.
SUMMARY
According to one embodiment of the present invention, a method for estimating a
Doppler spread and a signal-to-noise ratio of a received signal includes: (a) calculating one or more functions of a carrier component of the received signal at a first time point relative to a pilot signal embedded in the received signal; (b) Calculating the one or more functions of the carrier component of the received signal at a second time point relative to the pilot signal; (c) repeating steps (a) and (b) over multiple time periods, each time period being longer than the second time, and accumulating the one or more functions for the first time points and the second time points; and (d) Deriving the Doppler spread and the signal-to-noise ration based on the accumulated one or more functions calculated at the first and second time points, hi one embodiment, the first time point is substantially closer to the pilot signal than the second time point. The first time point is one symbol interval from the pilot signal, hi one embodiment, a product of the Doppler spread and the time difference between the first time point and the second time is approximately 0.25.
According to one embodiment of the present invention, the one or more functions may be an even-symmetric linearity function of a sum of the carrier component and the pilot signal. The one or more functions may also be an even-symmetric linearity of a difference between the carrier component and the pilot signal. The even-symmetric linearity may be an absolute value function, or a squaring function. The time domain representation of the received signal may be a sum of the transmitted signal and noise, a Rayleigh faded continuous wave signal or the Jake's model.
The present invention provides a method for estimating simultaneously Doppler spread and signal-to-noise ratio (SNR) of a received signal. According to one embodiment of the present invention, the method computes a loss in signal coherence over two different values of time intervals.
The present invention is better understood upon consideration of the detailed description below and the accompanying drawings
BRIEF DESCRIPTION OF THE DRAWINGS
The present invention is illustrated by way of example, and not limitation, in the figures of the accompanying drawings, in which:
Figure 1 illustrates a process for determining the Doppler spread and the SNR of a DVB-T/H OFDM signal, in accordance with embodiment of the present invention.
Figure 2 shows the locations of symbol intervals in a measurement cycle used in this example.
DETAILED DESCRIPTION
The present invention provides methods for simultaneously estimating the Doppler spread and SNR of a received signal by computing the loss in signal coherence over two different values of time intervals. Suitable signals for the present method includes a DVB- T/H OFDM signal, as defined in "ETSI EN 300 744, vl.5.1, Digital Video Broadcasting (DVB); Framing structure, channel coding and modulation for digital terrestrial television, European Telecommunications Standards Institute, 2004-06." While the present methods are applicable to any communication system that transmits a known signal such as a training sequence or a pilot subcarrier, a particular application to DVB-T/H OFDM signals is provided in this detailed description to illustrate the present methods. The Doppler spread and SNR may be estimated simultaneously to select optimum values of reception and demodulation parameters of a receiver to accommodate channel conditions change. The methods of the present invention avoid the cost of computing a number of complex- valued autocorrelations by simply computing the amplitude of autocorrelation at two different time offsets and jointly solving for the received SNR and the bandwidth of the Doppler spread.
Algorithms discussed in this detailed description are self-consistent sequences of steps which may represent physical manipulations of physical quantities of electrical or magnetic signals in the form of bits, values, elements, symbols, characters, terms, or numbers.
One embodiment of the present invention is illustrated in the following. Often, a mobile communications receiver receives a signal that has arrived from a single transmitter via multiple propagation paths. Each path may represent reflections of the signal by one or more reflectors. The combined sum of the many signals arriving along independent paths may exhibit Rayleigh fading. A Doppler effect may also result from the relative motions between the reflectors, the receiver, and the transmitter. When the path lengths vary independently, a different Doppler effect may affect the different signal components. Consequently, the combined sum of all signal components may exhibit a continuum of Doppler effects, referred to as a "Doppler spread."
According to the well-known Jakes' model, a Rayleigh faded continuous wave (CW) signal has a power spectrum, given by:
where fp is the maximum Doppler shift, f c is the frequency of the CW signal, and A is the mean power of the faded signal. The corresponding autocorrelation function is given by:
where Jo is the 0 th order Bessel function of the first kind.
If the Doppler spread is known, the receiver's demodulation and decoding performance may be optimized to reduce communication errors. For example, the channel estimation procedure for OFDM reception may be improved based on the Doppler spread spectrum. Similarly, the equalizer adaptation bandwidth for single-carrier reception may also be adjusted according to the Doppler spread. The SNR may also be used in the receiver to adjust response times and to properly weight information used in decoding decisions.
In accordance with one embodiment of the present invention, the SNR is estimated from the signal and noise powers of the subcarriers with known modulation. For a DVB-T/H OFDM signal, the continuous pilots may be used for this purpose. In a single-channel system (e.g., UMTS WCDMA), a pilot channel may be used for this purpose. After removing the known modulation, the phase of each subcarrier is continuous, and symbol pairs can be coherently summed and differenced to determine the signal and noise powers. As the time interval between the pair of symbols is increased, the signal loses phase coherence due to Doppler spread, which affects the signal and noise powers. The power measurements at two different time intervals between symbol pairs may then be used to determine jointly both the SNR and Doppler spread.
Figure 1 illustrates a process for determining the Doppler spread and the SNR of a
DVB-T/H OFDM signal, in accordance with embodiment of the present invention. Figure 2 shows the locations of symbol intervals in a measurement cycle used in this example As shown in Figure 1, at step 101, the received signal is down-converted from a radio frequency, sampled, and transformed by an fast fourier transform (FFT) to recover the OFDM subcarriers. In a DVB-T/H signal, Np of the subcarriers are continual pilots, which are each modulated to a known phase modulation in each symbol interval. The continual- pilot subcarriers are processed serially. At N c subsequent measurement cycles, at symbol
interval to, the complex subcarrier values of the N p
continual pilots are stored (step 102). At symbol intervals ti and tfc of each measurement cycle, the subcarriers are summed and differenced with the subcarrier values that are stored from symbol interval t 0
(steps 103 and 104, respectively) and accumulated (steps 105 and 106). In this embodiment, t 0
and U are adjacent symbol intervals. The powers of the sum and differences may be computed from a squaring operation, or any even-symmetric nonlinearity (e.g., an absolute value). The accumulated sums and differences are provided by: N sum )r n
(t 0
+ mT) + T n
(J 1
+ m7%
+mT)-r n
(t l
+mT)\
where r n (t) is the complex-valued FFT output for the n th subcarrier, and T is the measurement cycle longer than k+1 symbol intervals, as shown in Figure 2.
Each subcarrier output r n (t) may be modeled by the sum of a continual pilot s n (t) and noise n n (t) :
r n (t) = s n {t) + n n {t)
Thus, the mean values of the sum and difference measurements are
E{sum k ) = E(Js n (t 0 ) + /I 11 (0 + S n (t k ) + n n (t)f
= 2S(l + J 0 (2πf D τ k )) + 2N
E(diff k
) = - S n
(t k
) - n n
(0|
= 2S(l -|J 0 (2^r,)|) + 2N
where S is the mean power of the subcarrier signal, N is the power of the additive noise, and τ k is the correlation lag between to and t \ .
The mean values of sunn , diffi, sum^, and diff^ may be arithmetically combined in numerous ways to extract information about the Doppler spread/o and signal-to-noise ratio S/N. In one embodiment, the Doppler spread is determined from the ratio:
sum k - diff k 45 λWλ)! sum λ -diff λ = 4S λ ( 2 ¥O τ i )|
which can be then be used to numerically solved for the Doppler spread/b- The computed error is low when T 1 is one symbol interval and τ k is chosen such that the product f D x τ k is approximately 0.25. (Actually, any positive value may be used in theory; however, values less than unity are preferred.) Similarly, the SNR value S/N may be determined from the ratio:
Thus, methods for simultaneously estimating the Doppler spread and SNR of a received signal (e.g., a DVB-T/H OFDM signal) using the loss in signal coherence over two different values of time intervals have been described.
The above detailed description is provided to illustrate the specific embodiments of the present invention and is not intended to be limiting. Numerous variations and modifications within the scope of the present invention are possible. The present invention is set forth in the following claims.