Stochasticity in Feedback Loops: Great Expectations and Guaranteed Ruin

Stochastic feedback systems give rise to a variety of notions of stability. The conditions for the stability of the median, mean, and variance stability conditions differ. These conditions can be stated explicitly for scalar discrete-time systems with (almost) arbitrary distributions of the stochastic feedback gain. The state variable in such systems evolves towards a heavy-tailed distribution and exhibits some non-intuitive characteristics. For example, one can use stochastic feedback to stabilise unstable systems where one does not know the sign of the unstable pole or the sign of the system gain. A more dramatic example is an investment scheme which simultaneously yields unbounded expected profit and almost certain bankruptcy to every investor.

[1]  N. Elia.  Emergence of Power Laws in Networked Control Systems , 2006, Proceedings of the 45th IEEE Conference on Decision and Control.

[2]  W. Harkness,et al.  Generalized hyperbolic secant distributions , 1968 .

[3]  E. Hoversten,et al.  Communication theory for the turbulent atmosphere , 1970 .

[4]  C. Goldie.  IMPLICIT RENEWAL THEORY AND TAILS OF SOLUTIONS OF RANDOM EQUATIONS , 1991 .

[5]  G. D. Shellard.  Estimating the Product of Several Random Variables , 1952 .

[6]  Wilfred Perks,et al.  On Some Experiments in the Graduation of Mortality Statistics , 1932 .

[7]  Subhrakanti Dey,et al.  Heavy-tails in Kalman filtering with packet losses: confidence bounds vs second moment stability , 2018, 2018 European Control Conference (ECC).

[8]  Thomas Mikosch,et al.  Stochastic Models with Power-Law Tails: The Equation X = Ax + B , 2016 .

[9]  Bruno Sinopoli,et al.  Kalman filtering with intermittent observations , 2004, IEEE Transactions on Automatic Control.

[10]  Erik I. Verriest,et al.  Stability and stabilization of discrete systems with multiplicative noise , 1997, 1997 European Control Conference (ECC).

[11]  D. J. Finney,et al.  The Lognormal Distribution (With Special Reference to Its Use in Economics) , 1957 .

[12]  J. Clifton Samuels,et al.  On the mean square stability of random linear systems , 1959, IRE Trans. Inf. Theory.

[13]  J. Willems,et al.  Frequency domain stability criteria for stochastic systems , 1971 .

[14]  V. I. Oseledec.  A multiplicative ergodic theorem: Lyapunov characteristic num-bers for dynamical systems , 1968 .

[15]  Peter Hall,et al.  On the Limiting Behaviour of the Mode and Median of a Sum of Independent Random Variables , 1980 .

[16]  S. Ethier.  The Kelly system maximizes median fortune , 2004 .

[17]  H. Kesten.  Random difference equations and Renewal theory for products of random matrices , 1973 .

[18]  S. Bittanti,et al.  Periodic Systems: Filtering and Control , 2008 .

[19]  P. Spanos,et al.  Stochastic averaging: An approximate method of solving random vibration problems , 1986 .

[20]  Nicola Elia,et al.  Remote stabilization over fading channels , 2005, Syst. Control. Lett..

[21]  R. Fisher.  014: On the "Probable Error" of a Coefficient of Correlation Deduced from a Small Sample. , 1921 .

[22]  G. Blankenship,et al.  Stability of linear differential equations with random coefficients , 1977 .

[23]  Jacques L. Willems,et al.  Mean square stability criteria for stochastic feedback systems , 1973 .

[24]  Leo A. Goodman,et al.  The Variance of the Product of K Random Variables , 1962 .

[25]  D. J. Finney.  On the Distribution of a Variate Whose Logarithm is Normally Distributed , 1941 .

[26]  F. P. Cantelli.  Sui confini della probabilità , 1929 .

[27]  Richard L. Tweedie,et al.  Markov Chains and Stochastic Stability , 1993, Communications and Control Engineering Series.

[28]  A. Brandt.  The stochastic equation Yn+1=AnYn+Bn with stationary coefficients , 1986 .

[29]  Omar Fawzi,et al.  Bounds on Lyapunov Exponents via Entropy Accumulation , 2019, IEEE Transactions on Information Theory.

[30]  Laurence B. Milstein,et al.  Chernoff-Type Bounds for the Gaussian Error Function , 2011, IEEE Transactions on Communications.

[31]  Karl Johan Åström,et al.  On a first-order stochastic differential equation , 1965 .

[32]  Bruno Sinopoli,et al.  Kalman Filtering With Intermittent Observations: Tail Distribution and Critical Value , 2012, IEEE Transactions on Automatic Control.

[33]  Anders Rantzer,et al.  Concentration Bounds for Single Parameter Adaptive Control , 2018, 2018 Annual American Control Conference (ACC).

[34]  Hans Crauel,et al.  Stabilization of Linear Systems by Noise , 1983 .

[35]  B. Delyon,et al.  Concentration Inequalities for Sums and Martingales , 2015 .

[36]  Benoîte de Saporta,et al.  Tail of the stationary solution of the stochastic equation Yn+1=anYn+bn with Markovian coefficients , 2004 .

[37]  Jing Wang,et al.  Distributed Averaging Under Constraints on Information Exchange: Emergence of Lévy Flights , 2012, IEEE Transactions on Automatic Control.

[38]  R. Jungers,et al.  Lower and upper bounds for the largest Lyapunov exponent of matrices , 2013 .

[39]  Edward L. Dodd.  The Frequency Law of a Function of Variables With Given Frequency Laws , 1925 .

[40]  Peng Ding,et al.  Three Occurrences of the Hyperbolic-Secant Distribution , 2014, 1401.1267.

[41]  Stephen P. Boyd,et al.  Linear Matrix Inequalities in Systems and Control Theory , 1994 .

[42]  Matthias J. J. Fischer.  Generalized Hyperbolic Secant Distributions: With Applications to Finance , 2013 .