IEEE TRANSACTIONS ON INFORMATION THEORY, VOL.
IT-22,
NO. 1, JANUARY
[6] K. Yao, “A representation theorem and its applications to
spherically-invariant random processes,” IEEE Trans. Inform.
Theory,, vol. IT-19, pp. 600-608, Sept. 1973.
[7] B. Picmbono, “Spherically invariant and compound Gaussian
stochastic processes,” IEEE Trans. Inform. Theory (Corresp.),
vol. IT-16, pp. 77-79, Jan. 1970.
[8] B. Picinbono and G. Vezzosi,
“Detection d’un signal certain
dans un bruit non stationnaire et non Gaussien,” Ann. Tele-
commun., vol. 25, pp. 433-439, 1970.
[9] --, “Detection d’un signal certain dans un bruit spheriquement,
invariant, structure et caracteristiques des recepteurs,” Ann.
Telecommun., vol. 27, pp. 95-110, 1972.
[lo] D. K. McGraw and J. F. Wagner, “Elliptically symmetric
distributions,”
IEEE Trans. Inform. Theory, vol. IT-14, pp.
110-120, Jan. 1968.
[ll] J. Goldman, “Statistical properties of a sum of sinusoids and
Gaussian noise and its generalization to higher dimensions,”
Bell Syst. Tech. J., vol. 53, pp. 557-580, Apr. 1974.
[12] K. C. Chu, “Estimation and decision for linear systems with
elliptical random processes,” IEEE Trans. Automat. Contr.,
vol. AC-18, pp. 499-505, Oct. 1973.
[13] K. Yao, “Mutual information of spherically-invariant random
vectors and processes,” submitted to SIAM J. Appl. Math.
[14] D. Kelker, “Distribution theory of spherical distributions and a
location-scale parameter generalization,” Sankhya: Ind. J. Stat.,
ser. A, vol. 32, pp. 419-430, 1970.
1976
59
[15] R. Ahmad, “Extension of the normal theory to spherical families,”
Trabajos de Estadistica y de Investigation Operativa, vol. 23,
pp. 51-60, 1972.
[16] S. Bochner, Lectures on Fourier Integrals. Princeton, N.J.:
Princeton Univ. Press, 1959.
[17] K. S. Miller, Multidimensional Gaussian Distributions. New
York: Wiley, 1964.
[18] H. L. VanTrees, Detection, Estimation, and Modulation Theory,
Part I. New York: Wiley, 1968.
[19] W. Magnus, F. Oberhettmger, and R. P. Soni, Formulas and
Theorems for the Special Functions of Mathematical Physics.
New York: Springer, 1966.
[20] J. Goldman? “Multiple error performance of PSK systems with
cochannel interference and noise,” IEEE Trans. Commun.
Technol., vol. COM-19, pp. 420430, Aug. 1971.
[21] M. J. Wilmut and L. L. Campbell, “Signal detection in the
presence of cochannel interference and noise,” IEEE Trans.
Commun. Technol., vol. COM-20, pp. 115331158, Dec. 1972.
[22] V. K. Prabhu, “Error rate considerations for coherent phase shift
keyed systems with cochannel interference,” Bell Syst. Tech. J.,
vol. 48, pp. 743-767, Mar. 1969.
[23] M. Onoe, Tables of Modified Quotients of Bessel Functions of the
First Kind for Real and Imaginary Arguments. New York:
Columbia Univ. Press, 1958.
[24] I. S. Gradshteyn and I. M. Ryzhik, Tables of Integrals, Series,
and Products. New York: Academic, 1965.
Quantum-Mechanical Linear Filtering of
Random Signal Sequences
JOHN S. BARAS,
MEMBER, IEEE,
ROBERT 0. HARGER,
MEMBER, IEEE, AND
YOUNG H. PARK
Abstrucf-The problem of estimating a member of a scalar random
signal sequence with quantum-mechanical measurements is considered.
The minimum variance linear estimator based on an optimal present
quantum measurement and optimal linear processing of past measure-
ments is found. When the average optimal measurement without post-
processing, for a fixed signal, is linear in the random signal and the signal
sequence is pairwise Gaussian, the optimal processing separates: the
optimal measurement is the same as the optimal measurement without
regard to past data, and the past and present data are processed classi-
cally. The results are illustrated by considering the estimator of the
real amplitude of a laser signal received in a single-mode cavity along
with thermal noise; when the random signal sequence satisfies a linear
recursion, the estimate can be computed recursively. For a one-step
memory signal sequence it is shown that the optimal observable generally
differs from the optimal observable disregarding the past; the optimal
measurement can be computed recursively.
Manuscript received February 7, 1975; revised August 1, 1975.
This work was supported in part by the National Science Foundation
under Grant GK-14920.
J. S. Baras and R. 0. Harger are with the Electrical Engineering
Department, University of Maryland, College Park,, Md. 20742.
Y. H. Park is with Bendix Corporation, Columbta, Md.
I. INTRODUCTION
D
ETECTION and estimation problems have recently
been studied [ l]-[3] employing measurement models
correctly incorporating quantum mechanics. Such work
applies directly, e.g., to establishing fundamental limita-
tions in optical communication systems [4]. More recently,
the analog of filtering a random signal sequence has been
considered [S], [6], [13]. H
ere the problem of estimating
x,, a member of a “signal” sequence {x0,x1,. . . ,xk,. . .} of
scalar random variables, is considered; the parameter
k
is
conveniently regarded as discrete time. To be chosen are
the optimal measurements at time
k
and the optimal linear
combination of present and past measurements at times
j = 0, l;..,
k
- 1. The random sequence so obtained is
defined precisely below, but it is simply described in the
optical communication setting as follows.
At time
k
a laser signal modulated in some fashion by
xk is received in a cavity containing otherwise only an
electromagnetic field due to thermal noise; the total field is