Skip to main content
Proceedings of the National Academy of Sciences of the United States of America logoLink to Proceedings of the National Academy of Sciences of the United States of America
. 1998 Apr 28;95(9):4800–4803. doi: 10.1073/pnas.95.9.4800

Line spectral analysis for harmonizable processes

Keh-Shin Lii *, Murray Rosenblatt †,
PMCID: PMC20166  PMID: 9560181

Abstract

Harmonizable processes with spectral mass concentrated on a number of straight lines are considered. The asymptotic behavior of the bias and covariance of a number of spectral estimates is described. The results generalize those obtained for periodic and almost periodic processes.

Keywords: nonstationarity, spectral estimation, asymptotic bias and covariance


Let {Xt} be a continuous time parameter harmonizable process continuous in mean square, −∞ < t < ∞, with EXt ≡ 0. By this we mean that the covariance function r(t, τ) = E(XtInline graphicτ) has a Fourier representation

graphic file with name M2.gif 1

with F(λ, μ) a function of bounded variation. This implies that Xt itself has a Fourier representation in mean square

graphic file with name M3.gif 2

in terms of a random function Z(λ) with

graphic file with name M4.gif 3

If the process Xt is real-valued,

graphic file with name M5.gif 4

In the case of a weakly stationary process r(t, τ) = r(t − τ, 0) and all the spectral mass is located on the diagonal line λ = μ. If the process is periodic with

graphic file with name M6.gif

for some period a or almost periodic the spectral mass is located on a finite or countable number of lines in the (λ, μ) plane with slope one. If the process is discretely observed Xn, n = 0, ±1, ±2, … there is an analogous representation

graphic file with name M7.gif 5
graphic file with name M8.gif

with

graphic file with name M9.gif
graphic file with name M10.gif 6
graphic file with name M11.gif

The folding of F(λ, μ) to obtain H(λ, μ) is referred to as aliasing.

There is an enormous literature concerned with spectral estimation in the case of stationary processes (1). Recently efforts have been made to obtain analogous results on spectral estimation for periodic and almost periodic processes (27). It is well known that one generally does not have consistent estimates of spectral mass for a harmonizable process when the function F (or H) is absolutely continuous with a spectral density function f, dF(λ, μ) = f(λ, μ)dλdμ, with f(λ, μ) ≠ 0 on a set of positive two-dimensional Lebesgue measure, and one is sampling from the process Xn, … , Xn and n → ∞. A simple example is given by X0 normal with mean zero and variance one and Xk ≡ 0 with probability one for k ≠ 0. It is clear that consistency of spectral estimates in the case of stationary and periodic processes is due to the fact that the spectral mass is concentrated on lines that in these cases happen to be of slope one. We shall consider spectral estimation for harmonizable processes when the spectrum is concentrated on a finite (or possibly countable) number of lines. For convenience the slope of the lines will be assumed to be positive, though the modification for negative slopes is clear. A simple example of a harmonizable process with spectral mass on lines is given by

graphic file with name M12.gif 7

where Yt is stationary and βs and αs are real and positive numbers, respectively. The object is to give some insight into an interesting class of nonstationary processes.

Assume that Xt is a harmonizable real-valued process as already specified with

A1.  All its spectral mass on a finite number of lines with positive slope

graphic file with name M13.gif

A2.  The spectral mass on the line u = aiv + bi is given by a continuously differentiable spectral density fai,bi(v) if ai ≤ 1. Notice that the real-valued property implies that if u = av + b is a line of spectral mass, then so are the lines u = avb and u = a−1v ± a−1b. If there are lines of nonzero spectral mass, the diagonal λ = μ must be one of them with positive spectral mass. The condition

graphic file with name M14.gif

implies if u = av + b is a line of nonzero spectral mass, then so is u = a−1(vb) with

graphic file with name M15.gif

A3.  The spectral densities fai,bi(v) and their derivatives are bounded in absolute value by a function h(v) that is a monotonic decreasing function of |v| that decreases to zero as |v| → ∞ and that is integrable as a function of v.

As already remarked, aliasing or folding of the spectral mass occurs when the process is discretely sampled at times n = …  , − 1, 0, 1, … rather than continuously. The following simple remark indicates how a process with line spectra may differ from a stationary or almost periodic process in terms of aliasing. The aliasing in the case of a harmonizable process has a more complicated character.

Proposition 1. Let Xt be a continuous time parameter process continuous in mean square satisfying conditions A1–A3. Assume that the lines of spectral support have spectral density nonzero at all points v, |v| > s for some s > 0. The process discretely observed Xn then has a countably dense set of lines of support in [−π, π]2 if and only if one of the lines of spectral support of Xt has irrational slope a.

The Periodogram

We shall consider spectral estimation for the discretely observed process Xn. The estimates will be obtained by smoothing a version of the periodogram. Before dealing with the spectral estimates, approximations for the mean and covariances of the periodogram are obtained. Let

graphic file with name M16.gif 8

be the finite Fourier transform of the data xn/2, … , xn/2. The periodogram

graphic file with name M17.gif 9

and it is to be understood that |λ|, |μ| ≤ π with −π identified with π so that in effect one is dealing with the torus in (λ, μ). Set

graphic file with name M18.gif
graphic file with name M19.gif

These expressions are versions of the Dirichlet kernel adapted to (−π, π] and (−∞, ∞). The following result is useful in obtaining the expressions for the mean and covariance of the periodogram.

Lemma 1. If a > 0, |y| < π,

graphic file with name M20.gif 10

and

graphic file with name M21.gif
graphic file with name M22.gif
graphic file with name M23.gif 11

if |y| ≤ a/3, while if |y| ≥ a/3 the expression is itself of order log n/n.

Whenever we refer to an expression ω = z mod 2π it is understood that −π < ω ≤ π. Let {u} be the integer ℓ such that −1/2 < u − ℓ ≤ 1/2. Our version of z mod 2π is then z mod 2π = z − {z/(2π)}2π. In the following an approximation is given for EIn(αμ + ω, μ) with the condition imposed that α > 0 and −π < αμ + ω, μ ≤ π. Let y = y(k, a, b) = (2πka + (a − α)μ + b − ω) mod 2π.

Theorem 1. The mean

graphic file with name M24.gif 12
graphic file with name M25.gif

where in the sum it is understood that the k are integers and the pairs (a, b) correspond to the lines u = av + b in the spectrum of the continuous time parameter process Xt that one is observing at integer t.

In the following result an approximation is given for the cov(In(αμ + ω, μ), In(α′μ′ + ω′, μ′)). Here

graphic file with name M26.gif
graphic file with name M27.gif
graphic file with name M28.gif
graphic file with name M29.gif 13

with λ = αμ + ω, λ′ = α′μ′ + ω′, −π < λ, λ′, μ, μ′ ≤ π. Also (a, b), (a′, b′) correspond to lines in the spectrum of the continuous time parameter process Xt that one is observing at integer times.

Theorem 2. The covariance in the case of a normal process Xt is

graphic file with name M30.gif 14
graphic file with name M31.gif
graphic file with name M32.gif
graphic file with name M33.gif
graphic file with name M34.gif

Corollary 1. The result of Theorem 2 holds with an additional error term O(1/n) for a nongaussian harmonizable process with finite fourth-order moments if the fourth-order cumulants satisfy

graphic file with name M35.gif 15

This will be the case if

graphic file with name M36.gif 16

Spectral Estimates

We consider an estimate f̂α,ω(η) of f̃α,ω(η) obtained by smoothing the periodogram. Let K(η) be a nonnegative bounded weight function of finite support with ∫ K(η)dy = 1. The weight function Kn(η) = bn−1 K(bn−1η) with bn ↓ 0 as n → ∞ and nbn → ∞. The weight functions Kn should be considered as functions on the circle (−π, π] with −π identified with π. The estimate

graphic file with name M37.gif 17

Proposition 2. If A2 is strengthened so that the spectral densities fai,bi(ν) are assumed to be twice continuously differentiable and K is symmetric, then

graphic file with name M38.gif

where

graphic file with name M39.gif

as n → ∞.

The asymptotic behavior of covariances is described in the following result.

Theorem 3. Let Xt be a continuous time parameter harmonizable process continuous in mean square satisfying assumptions A1–A3 and 16. Then

graphic file with name M40.gif
graphic file with name M41.gif
graphic file with name M42.gif
graphic file with name M43.gif
graphic file with name M44.gif
graphic file with name M45.gif
graphic file with name M46.gif
graphic file with name M47.gif

where the first sum Σ′ is over a, a′, b, b′, k, k such that

graphic file with name M48.gif

with j′ = 2πka′ + b′ − ((2πka′ + b′)mod 2π), while the second sum Σ" is over a, a′, b, b′, k, k such that

graphic file with name M49.gif

with

graphic file with name M50.gif
graphic file with name M51.gif

In the almost periodic case we have the following corollary.

Corollary 2. If the assumptions of Theorem 3 are satisfied with Xt almost periodic

graphic file with name M52.gif
graphic file with name M53.gif

where the first sum is over b, b′, k, k such that

graphic file with name M54.gif
graphic file with name M55.gif

with j′ = 2πk′ + b′ − ((b′)mod 2π) while the second sum is over b, b′, k, k with

graphic file with name M56.gif
graphic file with name M57.gif

with

graphic file with name M58.gif
graphic file with name M59.gif

On heuristic grounds one would expect to be able to estimate a spectral density localized on a piecewise smooth curve in the plane.

Acknowledgments

This research is partly supported by Office of Naval Research Grant N00014-92-J-1086.

References

  • 1.Yaglom A M. Correlation Theory of Stationary and Related Random Functions. New York: Springer; 1986. , Vols. I and II. [Google Scholar]
  • 2.Dandawate A, Giannakis G. IEEE Trans Inf Theory. 1994;40:67–84. [Google Scholar]
  • 3.Gardner W, Franks L. IEEE Trans Inf Theory. 1975;21:4–14. [Google Scholar]
  • 4.Hurd H. IEEE Trans Inf Theory. 1989;35:350–359. [Google Scholar]
  • 5.Hurd H, Gerr N. J Time Ser Anal. 1991;12:337–350. [Google Scholar]
  • 6.Leskow J, Weron A. Stat Prob Lett. 1992;15:299–304. [Google Scholar]
  • 7.Leskow J. Stochastic Processes Appl. 1994;52:351–360. [Google Scholar]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences

RESOURCES