1: %% ****** Start of file template.aps ****** %
2: %%
3: %%
4: %% This file is part of the APS files in the REVTeX 4 distribution.
5: %% Version 4.0 of REVTeX, August 2001
6: %%
7: %%
8: %% Copyright (c) 2001 The American Physical Society.
9: %%
10: %% See the REVTeX 4 README file for restrictions and more information.
11: %%
12: %
13: % This is a template for producing manuscripts for use with REVTEX 4.0
14: % Copy this file to another name and then work on that file.
15: % That way, you always have this original template file to use.
16: %
17: % Group addresses by affiliation; use superscriptaddress for long
18: % author lists, or if there are many overlapping affiliations.
19: % For Phys. Rev. appearance, change preprint to twocolumn.
20: % Choose pra, prb, prc, prd, pre, prl, prstab, or rmp for journal
21: % Add 'draft' option to mark overfull boxes with black boxes
22: % Add 'showpacs' option to make PACS codes appear
23: % Add 'showkeys' option to make keywords appear
24:
25:
26:
27: %\documentclass[aps,pre,twocolumn,groupedaddress,showpacs,showkeys]{revtex4}
28: %\documentclass[aps,prl,preprint,superscriptaddress]{revtex4}
29:
30:
31: %\documentclass[aps,chaos,preprint,groupedaddress,showpacs,showkeys]{revtex4}
32: \documentclass[aps,chaos,twocolumn,groupedaddress,showpacs,showkeys]{revtex4}
33:
34: \usepackage{graphicx}% Include figure files
35: \usepackage{dcolumn}% Align table columns on decimal point
36: \usepackage{bm}% bold math
37: \usepackage{amssymb}
38: %\usepackage[cp1251]{inputenc}
39: %\usepackage[english,russian]{babel}
40:
41: %\usepackage{srcltx}
42:
43:
44: % You should use BibTeX and apsrev.bst for references
45: % Choosing a journal automatically selects the correct APS
46: % BibTeX style file (bst file), so only uncomment the line
47: % below if necessary.
48: \bibliographystyle{apsrev}
49:
50:
51: \newcommand{\correct}[1]{\par\vskip 2mm \hrule\vskip 0.5mm\noindent%
52: \textsl{#1}\par\vskip 0.5mm \hrule\vskip 2mm}
53: \renewcommand{\correct}[1]{#1}
54:
55: \begin{document}
56:
57:
58: %Title of paper
59: \title{Chaotic synchronization of coupled electron-wave systems with backward waves}
60:
61: \author{Alexander~E.~Hramov}
62: \email{aeh@cas.ssu.runnet.ru}
63: \author{Alexey~A.~Koronovskii}
64: \email{alkor@cas.ssu.runnet.ru}
65: \author{Pavel~V.~Popov}
66: \author{Irene~S.~Rempen}
67: \affiliation{Faculty of Nonlinear Processes, Saratov State
68: University, Astrakhanskaya, 83, Saratov, 410012, Russia}
69:
70: \date{\today}
71:
72:
73: \begin{abstract}
74: The chaotic synchronization of two electron-wave media with
75: interacting backward waves and cubic phase nonlinearity is
76: investigated in the paper. To detect the chaotic synchronization
77: regime we use a new approach, the so-called time scale
78: synchronization [Chaos, \textbf{14} (3) 603--610 (2004)]. This
79: approach is based on the consideration of the infinite set of
80: chaotic signals' phases introduced by means of continuous wavelet
81: transform. The complex space-time dynamics of the active media and
82: mechanisms of the time scale synchronization appearance are
83: considered.
84: \end{abstract}
85:
86: % insert suggested PACS numbers in braces on next line
87: \pacs{05.45.Xt, 05.45.Tp}
88:
89: \keywords{coupled oscillators, spatially extended system,
90: space-time chaos, chaotic synchronization, phase, wavelet
91: transform, microwave tubes}
92:
93: \maketitle
94:
95:
96:
97:
98: {\bf Synchronization of chaotic oscillators is one of the
99: fundamental phenomena of nonlinear dynamics. In this paper we
100: consider chaotic synchronization of two electron-wave media with
101: interacting backward waves and cubic phase nonlinearity. An
102: efficient approach to detecting of synchronous dynamics of
103: distributed systems is suggested based on the time scale
104: synchronization concept [Chaos, \textbf{14} (3) 603--610 (2004)].
105: The complex space-time dynamics of the active media and mechanisms
106: of the appearance of time scale synchronization are discussed.}
107:
108:
109: \section{Introduction}
110: \label{intro}
111:
112: In last decade the investigation of chaotic synchronization has
113: become a task of great interest \cite{Rosenblum:1996_PhaseSynchro,
114: Pikovsky:2000_SynchroReview, Boccaletti:2002_ChaosSynchro,
115: Pikovsky:2002_SynhroBook}. It can be observed in a lot of chaotic
116: oscillators of different nature, including physical and biological
117: systems \cite{Pikovsky:2002_SynhroBook,Ticos:2000_PlasmaDischarge,
118: Anishchenko:2000_humanSynchro, Tass:2003_NeuroSynchro,
119: Prokhorov:2003_HumanSynchroPRE}. Chaotic synchronization can be
120: used for secret signal
121: transmission~\cite{Murali:1993_SignalTransmission,
122: Chua:1997_Criptography, Yaowen:2002_ChaoticCommunication,
123: Garcia-Ojalvo:2001_ChaosCommunication}. It also seems very
124: important to study the appearance of chaotic synchronization in
125: distributed systems demonstrating space--time chaos
126: \cite{Kurths:1995_DistrSystems, Boccaletti:1999_ControllingChaos,
127: Bragard:2000, Ticos:2000_Plasma, Ahlers:2002, Valladares:2002,
128: Mendoza:2004}.
129: %
130: Four types of chaotic synchronization have been carried out for
131: chaotic systems with lumped parameters, namely phase
132: synchronization (PS), complete synchronization (CS), lag
133: synchronization (LS) and generalized synchronization (GS)
134: \cite{Pecora:1997_SynchroChaos, Pikovsky:2002_SynhroBook,
135: Pikovsky:2000_SynchroReview, Boccaletti:2002_ChaosSynchro}. On the
136: ranges of mentioned regimes the intermittence behaviour \cite
137: {Boccaletti:2000_IntermitLagSynchro} takes place.
138: %
139: To describe the phase synchronization the instantaneous phase
140: $\phi(t)$ of a chaotic continuous time series is usually
141: introduced~(see \cite{Rosenblum:1996_PhaseSynchro,
142: Pikovsky:2002_SynhroBook, Anishchenko:2002_SynchroEng,
143: Pikovsky:2000_SynchroReview}). The PS means the entrainment of
144: phases of chaotic signals, whereas their amplitudes remain chaotic
145: and uncorrelated. Simultaneously, if Fourier spectrum of a signal
146: $x(t)$ is complex (chaotic time series are characterized by
147: Fourier spectrum without the main frequency), it is not always
148: possible to describe the system dynamics with the help of one
149: instantaneous phase $\phi(t)$ (see
150: \cite{Anishchenko:2004_ChaosSynchro, Hramov:2004_Chaos} for
151: detail). In this case other methods can be applied to analyze the
152: non-phase coherent complex systems dynamics (e.g., via curvature
153: or via recurrence, see \cite{Osipov:2003_Synchro_Chaos,
154: Romano:2004_Phys.Lett.A} for detail).
155:
156: In our works \cite {Koronovskii:2004_JETPLettersEngl,
157: Hramov:2004_Chaos} the new approach to the analysis of the chaotic
158: synchronization is described, based on examination of continuous
159: set of phases $\phi_s(t)$ of chaotic time series introduced with
160: the help of the continuous wavelet transform \cite
161: {Daubechies:1992_WVTBook, Kaiser:1994_Wvt, alkor:2003_WVTBookEng}.
162:
163:
164: Let us consider continuous wavelet transform of time series $x(t)$
165: \begin{equation}
166: W(s,t_0)=\int\limits_{-\infty}^{+\infty}x(t)\psi^*_{s,t_0}(t)\,dt,
167: \label{eq:WvtTrans}
168: \end{equation}
169: where $\psi_{s,t_0}(t)$ is the wavelet--function related to the
170: mother--wavelet $\psi_{0}(t)$ as
171: \begin{equation}
172: \psi_{s,t_0}(t)=\frac{1}{\sqrt{s}}\psi\left(\frac{t-t_0}{s}\right).
173: \label{eq:Wvt}
174: \end{equation}
175: In equation (\ref{eq:WvtTrans}) $x(t)$ is the time series
176: generated by a dynamical system. It may also contain the noise
177: component. The time scale $s$ corresponds to the width of the
178: wavelet function $\psi_{s,t_0}(t)$, and $t_0$ is the shift of
179: wavelet along the time axis, symbol ``$*$'' in~(\ref{eq:WvtTrans})
180: denotes complex conjugation. It should be noted that the time
181: scale $s$ is usually used instead of the frequency $f$ of Fourier
182: transform and can be considered as the quantity inversed to it.
183:
184: The Morlet--wavelet~\cite{Grossman:1984_Morlet}
185: \begin{equation}
186: \psi_0(\eta)=\frac{1}{\sqrt[4]{\pi}}\exp(j\Omega_0\eta)\exp\left(\frac{-\eta^2}{2}\right)
187: \label{eq:Morlet}
188: \end{equation}
189: has been used as a mother--wavelet function. The choice of
190: parameter value $\Omega_0=2\pi$ provides the relation ${s=1/f}$
191: between the time scale $s$ of wavelet transform and frequency $f$
192: of Fourier transform.
193:
194: The wavelet surface
195: \begin{equation}
196: W(s,t_0)=|W(s,t_0)|e^{j\phi_s(t_0)} \label{eq:WVT_Phase}
197: \end{equation}
198: describes the system's dynamics on every time scale $s$ at moment
199: $t_0$. The value of $|W(s,t_0)|$ indicates the presence and
200: intensity of the time scale $s$ mode in the time series $x(t)$ at
201: the moment of time $t_0$. It is also possible to consider the
202: quantity
203: \begin{equation}
204: \langle E(s)\rangle=\int|W(s,t_0)|^2\,dt_0, \label{eq:IntEnergy}
205: \end{equation}
206: which is the distribution of integral energy by time scales.
207:
208: At the same time, the phase $\phi_s(t)=\arg\,W(s,t)$ is naturally
209: introduced for every time scale $s$. In other words, $\phi_s(t)$
210: is a continuous function of time $t$ and time scale $s$. It means
211: that it is possible to describe the behavior of each time scale
212: $s$ by means of its own phase $\phi_s(t)$. Let us consider the
213: dynamics of two coupled oscillators. If in the time series
214: $\mathbf{x}_{1,2}(t)$ generated by these systems there is a range
215: of time scales $s_1\leq s\leq s_2$ for which the phase locking
216: condition
217: \begin{equation}
218: |\phi_{s1}(t)-\phi_{s2}(t)|<\mathrm{const} \label{eq:PhaseLocking}
219: \end{equation}
220: is satisfied and the part of the wavelet spectrum energy in this
221: range is not equal to zero
222: \begin{equation}
223: E_{snhr}=\int\limits_{s_1}^{s_2}\langle E(s)\rangle\,ds>0,
224: \label{eq:SynchroEnergy}
225: \end{equation}
226: then we assert that \textit{time scale synchronization} (TSS)
227: between oscillators takes place \cite{Hramov:2004_Chaos}. The
228: condition (\ref{eq:PhaseLocking}) can be generalized to the case
229: of $m:n$ synchronization. In this case one has to use the
230: condition
231: \begin{equation}
232: |m\phi_{sn1}(t)-n\phi_{sm2}(t)|<\mathrm{const}
233: \label{eq:PhaseLockingMN}
234: \end{equation}
235: instead of (\ref{eq:PhaseLocking}). In this case the time scales
236: $s_{m1}$ in the first system and $s_{n2}$ in the second system
237: would satisfy the relation $s_{m2}/s_{n1}=m/n$.
238:
239:
240: In Reference \cite {Hramov:2004_Chaos} the new approach was
241: applied to the study of chaotic synchronization of coupled
242: R\"ossler systems and two Chua's systems. It was shown, that
243: different types of synchronous oscillations (PS, LS, ILS, CS and
244: GS) may be considered as particular cases of TSS. It seems rather
245: interesting to apply the concept of time scale synchronization to
246: the analysis of chaotic synchronization of the distributed systems
247: demonstrating space-time chaos.
248:
249: In this paper we present the results of investigation of chaotic
250: synchronization in the coupled electron--wave systems with
251: backward electromagnetic wave and cubic nonlinearity, derived with
252: the help of the method of time scale synchronization.
253: %
254: The explored system is a simple model of microwave oscillator with
255: backward wave (BWO) finding wide practical application (see, for
256: example, \cite{Soohoo:1971_MicrowaveElectronicsBook,
257: book:1983_microwave, Granatstein:1987_Book,
258: Felch:1999_SmallAuthor}). For similar systems the complex chaotic
259: regimes of space--time oscillations are typical
260: \cite{Bezruchko:1979_BWOChaos, Ginzburg:1979_engl,
261: Levush:1992_BWO, Trubetskov:1996_CHAOS, Nusinovich:2001,
262: Dronov:2004_TWTChaos}.
263:
264:
265: The structure of the paper is as follows.
266: %
267: Section~\ref{Sct:BaseEq} contains the formalism describing
268: nonlinear interaction of electron beam and backward
269: electromagnetic wave.
270: %
271: Section~\ref{Sct:Synchro} describes the synchronization of two
272: unidirectional coupled electron-wave systems. We demonstrate the
273: efficiency of TSS method for the analysis of synchronization of
274: spatially extended systems and discuss the characteristics of
275: chaotic synchronization of such systems.
276: %
277: The quantitative measure of synchronization of spatially extended
278: systems is described in section~\ref{Sct:Measure}. In
279: section~\ref{Sct:SpaceDynamics} we consider complex space-time
280: dynamics of the unidirectional coupled electron-wave systems with
281: the help of the concept of time-scale synchronization.
282: %
283: The final conclusion is presented in section~\ref{Sct:Conclusion}.
284:
285:
286:
287:
288: \section{General formalism}
289: \label{Sct:BaseEq}
290:
291: Let us consider a simple model of two bound electron--wave media
292: with backward wave and cubic phase nonlinearity.
293:
294:
295: One of the types of the active media of electron--wave nature is
296: based on resonance interaction of electron wave and backward
297: electromagnetic wave in the slow-wave structure
298: \cite{Soohoo:1971_MicrowaveElectronicsBook, book:1983_microwave,
299: Granatstein:1987_Book, TrueAeh:2003_microwave_electronics_1engl}.
300: The electron wave is really a perturbation of density and velocity
301: of charged particles of electron beam propagating with the
302: velocity $v_0$ along a slow-wave structure.
303:
304: In the linear approximation such system conforms to a model of
305: interaction of two opposing linear waves with dispersion relation
306: \begin{equation}
307: \omega_1=k_1, \qquad \omega_2=-k_2.
308: \end{equation}
309: In this case the linear model can be formulated as a
310: self-consistent system of two equations
311: \begin{equation}\label{ref_bwopp_01}
312: \frac{\partial F}{\partial \tau}-\frac{\partial F}{\partial \xi}=-{A}I,
313: \end{equation}
314: \begin{equation}\label{ref_bwopp_02}
315: \frac{\partial I}{\partial \tau}+\frac{\partial I}{\partial \xi}=-{A}F,
316: \end{equation}
317: where an absolute instability is observed. Here
318: $F=|F|\exp(j\varphi_F)$ and $I=|I|\exp(j\varphi_I)$ are the
319: normalized slowly variable amplitudes of interacting
320: electromagnetic and electron waves, respectively, $\tau$ is the
321: normalized time, $\xi$ is the normalized coordinate and $A$ is the
322: normalized interaction parameter, which is proportional to the
323: electron beam current and the length of the active medium.
324:
325:
326:
327: Solution of the linear equations (\ref{ref_bwopp_01}) and
328: (\ref{ref_bwopp_02}) predicts unlimited exponential growth of
329: amplitude of each of the waves. In this case it is necessary to
330: take into account the nonlinear effects. One of the fundamental
331: nonlinear effects is the non-isochronism of electrons-oscillators,
332: expressed in dependence of the frequency of the
333: electron-oscillator on its energy. Non-isochronism of
334: electrons-oscillators in the system of interacting waves is
335: displayed in the shift of nonlinear phase $\varphi_I$ of the
336: electron wave. Replacing the old variables in the equations
337: (\ref{ref_bwopp_01}) and (\ref{ref_bwopp_02}) to new ones: $\tau '
338: = (\tau-\xi)/2 $ and $\xi ' = \xi$ we can rewrite
339: (\ref{ref_bwopp_02}) as
340: \begin{equation}\label{ref_bwopp_2}
341: \frac{\partial
342: I}{\partial\xi}+j|I|\frac{\partial\varphi_I}{\partial\xi}=-A|F|\exp({j(\varphi_F-\varphi_I))},
343: \end{equation}
344: where primes above new variables are omitted.
345:
346:
347:
348: From the point of view of the non-isochronism effect the
349: expression for $\partial\varphi_I/\partial\xi $ can be obtained
350: from the following reasons. The equation for the complete phase
351: $\varphi_I$ of electron wave must contain the item
352: $({\omega_e/v_e})\xi$, defining the shift of the frequency of
353: radiation due to the Doppler effect (here $\omega_e$ is the
354: frequency of the electron wave and $v_e$ is the longitudinal
355: velocity of its propagation). So, when we take into account the
356: dependence of $\omega_e$ upon the energy of the particles ${\cal
357: W}$, the item proportional to energy ${\cal W}$ should be added at
358: first approximation into the expression for a slow phase:
359: \begin{equation}\label{ref_bwopp_3}
360: \omega_e=\omega_{e0}+\left({\frac{\partial\omega_e}{\partial{\cal W}}}\right)_0{\cal W}+\dots
361: \end{equation}
362:
363:
364:
365: Supposing, that non-perturbed frequency $\omega _ {e0}$ is taken
366: into account in the "fast" phase, and the energy $ {\cal W} $ of
367: electrons (in case when their trajectories are identical) is
368: defined by the energy of the wave by the expression ${\cal W} =
369: \alpha|I |^2$, where $\alpha$ is the coefficient of
370: proportionality, we shall discover, that
371: \begin{equation}\label{ref_bwopp_4}
372: \frac{\partial\varphi_I}{\partial\xi}=\frac{\partial\varphi'_I}{\partial\xi}+
373: \left(\frac{\partial\omega_e}{\partial{\cal W}}\right)_0{\cal W}\frac{\alpha|I|^2}{v_e}=
374: \frac{\partial\varphi'_I}{\partial\xi}+\mu|I|^2,
375: \end{equation}
376: where the derivative $\partial\varphi'_I/\partial\xi$ describes
377: the phase shift of the wave only due to the interaction between
378: electrons and electromagnetic wave $F$; $\mu=({\alpha}/{v_e})\left
379: ({{\partial\omega_e}/{\partial{\cal W}}}\right)_0 \! \! {\cal W}$
380: is the parameter characterizing the change of the electron
381: velocity agreeing with the change of the electron energy during
382: the interaction. We shall now define the amplitude of an electron
383: wave as $I=|I|\exp({j\varphi'_I})$ and deal with the new variable
384: $F'=F/\sqrt{\mu}$, $I'=I/\sqrt{\mu}$, so the equation
385: (\ref{ref_bwopp_2}) will be written as follows
386: \begin{equation}\label{ref_bwopp_5}
387: \frac{\partial I}{\partial\xi}+j|I|^2I=-AF,
388: \end{equation}
389: where the primes above new variable are omitted.
390:
391: Thus, the equation (\ref{ref_bwopp_01}) for the backward electromagnetic wave
392: will not change.
393:
394:
395: The equations~(\ref{ref_bwopp_01}) and (\ref{ref_bwopp_5})
396: describe the nonlinear oscillations in the microwave spatially
397: extended system of interacting backward waves with cubic phase
398: nonlinearity. Such electron--wave system, as wtitten above, is a
399: simple model of microwave oscillator with backward wave (BWO \cite
400: {Soohoo:1971_MicrowaveElectronicsBook, book:1983_microwave} and
401: gyrotron BWO \cite {book:1983_microwave, Felch:1999_SmallAuthor,
402: Nusinovich:2001}). This system is suitable for the analysis of
403: chaotic oscillations, including chaotic synchronization, in
404: microwave tubes (see, for example, BWO
405: \cite{Bezruchko:1979_BWOChaos, Ginzburg:1979_engl,
406: Levush:1992_BWO} and gyrotron backward wave generator
407: \cite{Nusinovich:2001, Felch:1999_SmallAuthor}).
408:
409: In an autonomous regime of BWO it is necessary to supplement the
410: equations~(\ref{ref_bwopp_01}) and (\ref{ref_bwopp_5}) with the
411: standard boundary conditions $F(\xi=1, \tau)=0$ and
412: $I(\xi=0,\tau)=0$ describing the lack of electromagnetic and
413: electron waves on the boundaries of the autonomous system. In the
414: autonomous microwave oscillator with backward wave at
415: $\pi/2<A<1.83$ the single-frequency oscillations with stationary
416: space distribution of electromagnetic field $F(\xi)$ and current
417: $I(\xi)$ are observed. At $A>1.83$ the multifrequency periodic
418: oscillations (regime of a periodic self-modulation of an output
419: field \cite{Ginzburg:1979_engl, Levush:1992_BWO, Nusinovich:2001})
420: take place. At $A>2.05 $ the time series of the output field
421: $F(\tau, \xi=0)$ look like pulses with fine oscillations between
422: them, as a result of excitation of the complex irregular space
423: distribution $F(\xi) $ and $I(\xi)$. The reason of the latter is
424: the fast shift of the nonlinear phase of the electron wave along
425: the space coordinate of the distributed system. At last, at
426: $A>4.5$ chaotic wide-band oscillations are observed.
427:
428: In present paper we consider the system of two unidirectionally
429: coupled, spatially extended electron-wave systems in the chaotic
430: regime. The system is described by the following equations
431: \begin{equation}\label{ref_bwopp_01_TWO}
432: \frac{\partial F_{1,2}}{\partial \tau}-\frac{\partial F_{1,2}}{\partial \xi}=-{A_{1,2}}I_{1,2},
433: \end{equation}
434: \begin{equation}\label{ref_bwopp_5_TWO}
435: \frac{\partial I_{1,2}}{\partial \xi}+j|I_{1,2}|^2I_{1,2}=-A_{1,2}F_{1,2},
436: \end{equation}
437: where index ``$1$'' corresponds to the drive system and index
438: ``$2$'' - to the response one.
439:
440: Unidirectional connection between microwave oscillators is brought
441: into the system as a non-stationary boundary condition for the
442: slowly varying amplitude of the electromagnetic field $F_2$ of the
443: response system thus the boundary condition for the first
444: conducting fissile medium remains constant
445: \begin{equation}\label{ref_bwopp_boundary_properties_TWO1}
446: I_1(\xi=0,\tau)=0,\quad F_1(\xi=1,\tau)=0,
447: \end{equation}
448: \begin{equation}\label{ref_bwopp_boundary_properties_TWO2}
449: I_2(\xi=0,\tau)=0,\quad F_2(\xi=1,\tau)= \rho F_1(\xi=0,\tau).
450: \end{equation}
451: Hence $\rho = R\exp[j\theta]$ is the complex coupling coefficient
452: ($R$ is the coupling strength and $\theta$ is the phase of
453: coupling coefficient).
454:
455: For drive and response systems the values of control parameters
456: are taken as $A_1=4.2 $ and $A_2=4.9$, that corresponds to the
457: regimes of chaotic oscillations in the active electron--wave
458: systems with backward waves. The highest Lyapunov exponents are
459: $\lambda_{A1}=0.224\pm0.010$ and $\lambda_{A2}=0.617 \pm 0.006$
460: for the first and second control parameter values, respectively.
461: The value of the phase of coupling coefficient $\theta$ does not
462: influence the processes in the coupled systems and is fixed
463: $\theta=\pi$. In present paper the establishment of the chaotic
464: synchronization regime in two spatially extended electron--wave
465: systems (BWO) is analyzed at the variation of coupling strength
466: $R$.
467:
468:
469:
470:
471: \section{Chaotic synchronization of spatially extended electron--wave systems}
472: \label{Sct:Synchro}
473:
474:
475: Let us consider the behavior of the coupled systems with the
476: increase of amplitude of coupling coefficient $R$ and the fixed
477: values of control parameters (the values mentioned above are
478: taken). We shall investigate the oscillations $F_{r1,2}(\tau)=
479: \textrm{Re}\,\left[{F_{1,2}(\xi=0,\tau)}\right]$ at outputs
480: $\xi=0$ of both coupled systems.
481:
482:
483: Figures~\ref{fgr:spectrums_R}{a} and~\ref{fgr:spectrums_R}{b} show
484: the power spectra and phase pictures (reconstructed by
485: delay--coordinate embedding method \cite{Packard:1980_Delay,
486: Takens:1981}) of oscillations at the outputs of drive and response
487: systems in autonomous regimes (i.e. at $R=0$).
488: %
489:
490: \begin{figure}
491: \centerline{\scalebox{0.45}{\includegraphics{Fig1.ps}}}
492: \caption{Power spectra and phase pictures of output
493: electromagnetic field oscillations. (a) Drive oscillator with the
494: value of dimensionless parameter $A=4.2$. (b) Response oscillator
495: without external signal ($A=4.9$ and $R=0$). (c,d) Response
496: oscillator with the external signal. The value of coupling
497: parameter $R$ is (c) $R=0.2$ and (d) $R=0.5$
498: \label{fgr:spectrums_R}}
499: \end{figure}
500:
501:
502: The output field dynamics $F_{r1}$ of the drive oscillator (which
503: acts upon the input of the interaction space $\xi=1$ of response
504: oscillator) is chaotic (see Fig.\,\ref{fgr:spectrums_R}{a}). In
505: the power spectrum the high noise background at the level
506: $-30\,$dB is observed.
507: %
508: There are two main spectral components with frequencies $f_b$ and
509: $f_m$ in this spectrum.
510: %
511: High frequency $f_b$ is close to frequency $\hat\omega/2\pi$ of
512: synchronism of the linear noninteracting electron and backward
513: electromagnetic waves.
514: %
515: At the decrease of dimensionless parameter of the electron beam
516: $A_1$ the regime of periodic oscillations with the frequency close
517: to $f_b$ is established.
518: %
519: The second frequency which is marked as $f_m$ and is settling down
520: in a low-frequency part of a power spectrum, determines the
521: low-frequency modulation of complex amplitude $|F_1(\tau)|$ of the
522: electromagnetic backward wave.
523:
524: With the growth of control parameter $A$ the oscillations in the
525: electron--wave system complicate, that is obviously seen from the
526: analysis of the power spectrum and the chaotic attractor in
527: Fig.\,\ref{fgr:spectrums_R}({b}), of the second system at $A=4.9$
528: and $R=0$ (autonomous oscillations).
529: %
530: It is shown in Fig.\,\ref{fgr:spectrums_R}({b}), that the noise
531: background of the power spectrum rises up to a level
532: $-(5\div10)$\,dB. With such noise background it is already
533: impossible to select the main spectral components.
534:
535: Let us consider the behavior of the response system under the
536: influence of the drive system signal, at the increase of coupling
537: coefficient $R$. The growth of the coupling strength parameter $R$
538: means that the power of external signal acting on the response
539: system, increases as $R^2$ .
540:
541: \begin{figure}
542: \centerline{\scalebox{0.35}{\includegraphics{Fig2.ps}}}
543: \caption{The dependence of the synchronized time scale range
544: $[s_1;s_2]$ on the coupling strength $R$ for base time scales
545: $s_b$ (bottom) and $s_m$ (top)
546: \label{fgr:ScaleDifference}}
547: \end{figure}
548:
549:
550: With the growth of a coupling parameter $R$ the power spectrum of
551: generation of the response system changes that is illustrated by
552: Fig.\,\ref{fgr:spectrums_R}({c,\,d}).
553: %
554: At the small $R<0.3$ [Fig.\,\ref{fgr:spectrums_R}(c)] upon the
555: noise base of the response system spectrum there is a weakly
556: delineated frequency (marked by an arrow), corresponding to the
557: base frequency $f_b $ of the driving system spectrum [see
558: Fig.\,\ref{fgr:spectrums_R}(a)]. At the large values $R>0.3$ the
559: energy of the spectral component corresponding to the base
560: frequency (marked by arrow in Fig.\,\ref{fgr:spectrums_R}(d);
561: $R=0.5$) increases.
562:
563: It is necessary to mark, that both systems have attractors with
564: ill-defined phase. For these chaotic signals it is impossible to
565: introduce the instantaneous phase $\phi(t)$ of chaotic signal
566: correctly. It is clear that in such case the traditional methods
567: of the phase synchronization detecting (see Ref.
568: \cite{Rosenblum:1996_PhaseSynchro, Anishchenko:2004_ChaosSynchro,
569: Anishchenko:2002_SynchroEng, Pikovsky:2000_SynchroReview}) fail.
570: On the contrary, the approach based on introduction of
571: instantaneous phase $\phi_s(t)$ by the wavelet transform
572: \cite{Koronovskii:2004_JETPLettersEngl, Hramov:2004_Chaos} can
573: easily help to detect the time scale synchronization between
574: chaotic oscillators.
575:
576:
577:
578: The dependence of synchronized time scale range $[s_1;s_2]$ on
579: coupling parameter $R$ is shown in Fig.~\ref{fgr:ScaleDifference}.
580: The appearance of the synchronized time scale range corresponds to
581: the TSS regime.
582:
583: The peculiarity of dynamics of the system of two unidirectional
584: coupled spatially extended systems is that the TSS of chaotic
585: oscillations of each of subsystems at a large value of coupling
586: parameter $R$ is observed for two ranges near to base time scales
587: $s_b=1/f_b $ and $s_m=1/s_m $, respectively. In
588: Fig.\,\ref{fgr:ScaleDifference} the ranges of phase locked scales
589: are shown for the areas near to the base time scale $s_b$ (lower
590: figure), and the time scale $s_m $ (upper figure).
591:
592: From Fig.\,\ref{fgr:ScaleDifference} it can be seen, that for
593: coupling coefficient values $R<0.29$ the establishment of a
594: chaotic synchronization regime is not observed. The latter means,
595: that there are no time scales $s $ which would satisfy the
596: conditions of TSS (\ref{eq:PhaseLocking}) and
597: (\ref{eq:SynchroEnergy}).
598:
599:
600:
601:
602: Figure~\ref{fgr:WaveletPhases} illustrates the dynamics of phase
603: difference $\phi_{s1}(t)-\phi_{s2}(t)$ for different time scales
604: for $R=0.2$. It is clear, that there is no time scale range
605: $[s_1;s_2]$ for which the phase difference $\phi_{s1}(t) -
606: \phi_{s2}(t)$ ($s\in[s_1;s_2]$) is bounded.
607:
608: \begin{figure}
609: \centerline{\scalebox{0.39}{\includegraphics{Fig3.ps}}}
610: \caption{The phase difference $\phi_{s1}(t) - \phi_{s2}(t)$ for
611: two unidirectionally coupled spatially extended electron--wave
612: systems. (a) $R=0.2$. There is no synchronization between systems.
613: (b) $R=0.3$. The time scales $s=1.055$ (red line) of both systems
614: are correlated with each other and the time scale synchronization
615: regime can be observed
616: \label{fgr:WaveletPhases}}
617: \end{figure}
618:
619:
620: For the coupling parameter $R$ exceeding the critical value
621: $R_c\simeq 0.29$ we observe the time scale range $\Delta
622: s=s_2-s_1$ for which the requirement of capture of phases
623: (\ref{eq:PhaseLocking}) is satisfied and the energy relating on
624: this range (\ref{eq:SynchroEnergy}) is nonzero. The latter means,
625: that in the coupled active media the TSS regime is established
626: \cite{Koronovskii:2004_JETPLettersEngl, Hramov:2004_Chaos}.
627:
628: The behavior of the phase difference $\phi_{s1}(t) - \phi_{s2}(t)$
629: for this case is presented in Fig.\,\ref{fgr:WaveletPhases}(b).
630: So, we can say, that the time scales $s=1.055$ (and close to them)
631: of chaotic oscillations of electromagnetic field amplitudes on the
632: output of coupled systems are synchronized with each other. At
633: the same time the other time scales (e.g., $s=1.5$) remain
634: uncorrelated. For such time scales the phase locking is not
635: observed [see Fig.~\ref{fgr:WaveletPhases}(b)].
636:
637: Let us note, that with the growth of coupling parameter the TSS
638: appears, at first, near the base scale $s_b$ at $R=R_b$ (it is
639: marked in Fig.\,\ref{fgr:ScaleDifference}). With the further
640: increase of coupling coefficient, at $R=R_m> R_b $ the phase
641: locking of time scales occurs in the area of a time scale $s_m$,
642: corresponding to modulating frequency $f_m$ of amplitude of the
643: output field of the response BWO.
644:
645:
646: Let us consider at first, why the range $\Delta s$ of synchronized
647: time scales (i.e. time scales, for which the conditions of
648: synchronization (\ref{eq:PhaseLocking}) and
649: (\ref{eq:SynchroEnergy}) are fulfilled) appears near the time
650: scale $s_b$ corresponding to the base frequency $f_b=1/s_b $.
651: %
652: For this purpose we shall study wavelet power spectra $ \langle {E
653: _ {1,2} (s)} \rangle $, defined from the equation
654: (\ref{eq:IntEnergy}), for drive and response systems.
655: %
656: Wavelet spectra of the output field oscillations of drive (a
657: stroke line~{\sl1}) and response (a solid line~{\sl2}) BWO are
658: presented in Fig.\,\ref{fgr:WaveletSpectra} for different values
659: of coupling parameter $R$.
660: %
661: Let us mark, that the wavelet power spectrum
662: $\langle{E_1(s)}\rangle$ of the drive system does not vary (since
663: the value of controlling parameter $A_1 $ is fixed and the system
664: is autonomous) and is presented in all the figures, for different
665: values of $R$, as a matter of convenience comparison with the
666: wavelet power spectra $\langle{E_2}\rangle$ of the response
667: system.
668:
669: \begin{figure}
670: \centerline{\scalebox{0.7}{\includegraphics{Fig4.ps}}}
671: \caption{The energy distribution in wavelet spectra $\langle
672: E(s)\rangle$ for output field oscillations of drive (stroke line)
673: and response (solid line) spatially extended systems. (a) $R=0$.
674: There are autonomous oscillations in the systems. (b) $R=0.2$.
675: There is no time scale synchronization between systems. (c)
676: $R=0.3$. There is time scale synchronization near the base time
677: scale $s_b$ (Synchronized scale region is marked by gray box $S$).
678: (d) $R=0.5$. Time scale synchronization near both base time scale
679: $s_b$ and modulation time scale $s_m$ is observed.
680: \label{fgr:WaveletSpectra}}
681: \end{figure}
682:
683:
684: In Fig.\,\ref{fgr:WaveletSpectra}({a}) wavelet power spectra of
685: autonomous oscillations are shown ($R=0$).
686: %
687: One can see that the oscillations in each of the spatially
688: extended systems have essentially different wavelet power spectra.
689: %
690: So, in wavelet spectrum $\langle{E_1(s)}\rangle$ of the drive
691: system the base time scales $s_b $ (corresponding to the high
692: frequency $f_b$) and $s_m$ (corresponding to modulation
693: oscillations with the frequency $f_m=1/s_m$) stand out against the
694: background. In wavelet spectrum of an autonomous response system
695: the time scales corresponding to high-frequency spectral component
696: (they are marked by an arrow in Fig.\,\ref{fgr:WaveletSpectra}(a))
697: also precisely stand out against background. However, the
698: arrangement of a peak of $s_{b}$ in wavelet power spectrum
699: $\langle {E_2(s)}\rangle $ does not correspond to a base time
700: scale $s_b$ of the drive system.
701: %
702: As to large time scales (corresponding to low modulation
703: frequencies), their wavelet power spectra are essentially
704: different.
705: %
706: The wavelet spectrum of the response system in the area of the
707: time scale $s_m$ is continuous and homogeneous, i.e. the
708: modulation oscillations are characterized by continuous noise-like
709: spectrum.
710: %
711: For small values of coupling coefficient $R$, the shape of wavelet
712: power spectrum changes [see Fig.\,\ref{fgr:WaveletSpectra}({b})).
713: However, time scale synchronization is not observed, which is
714: apparent from Fig.\,\ref{fgr:WaveletPhases}(a).
715:
716: With the increase of $R$ the conditions of TSS
717: (\ref{eq:PhaseLocking}) and (\ref{eq:SynchroEnergy}) are satisfied
718: only by the time scales $s$ whose energy in power spectra of the
719: drive and response systems is rather large. Really, at the values
720: of coupling coefficient $R\approx0.3 $ as it was already
721: considered above, the conditions of TSS are fulfilled only for
722: time scales near the basic scale $s_b=1/f_b\approx1.0 $ (see also
723: Fig.\,\ref{fgr:ScaleDifference}).
724: %
725: As it follows from Fig.\,\ref{fgr:WaveletSpectra}(c), in the range
726: of time scales marked in Fig.\,\ref{fgr:WaveletSpectra}(c) by grey
727: color and symbol $S$, the energy of wavelet power spectrum is
728: maximum.
729:
730: From Fig.\,\ref{fgr:WaveletSpectra}(c) one can also see that the
731: energy of wavelet spectrum, being fallen on the time scale $s_m $
732: of the response system simultaneously increases. However, at
733: $R\approx0.3$ its energy is not large enough, and TSS of field
734: modulation oscillations is not observed.
735:
736: At large coupling coefficients $R> 0.35 $ the shape of wavelet
737: power spectra of output field oscillations of response and drive
738: systems become close to each other (see Fig.\,\ref
739: {fgr:WaveletSpectra}(d)).
740: %
741: Base time scales $s_m$ and $s_b$ are presented in the power
742: spectra with approximately equal intensity. The analysis of time
743: scales dynamics shows, that the conditions of synchronization
744: (\ref{eq:PhaseLocking}) and (\ref{eq:SynchroEnergy}) are fulfilled
745: for both most intensive time scales in wavelet power spectra and
746: two ranges of time scales, in which phase locking occurs [see
747: Fig.\,\ref{fgr:ScaleDifference} for $R> 0.35$], are accordingly
748: selected.
749:
750: Thus, the TSS of two coupled chaotic spatially extended systems is
751: observed first of all for those time scales $s$ whose energy in
752: the chaotic power spectrum is great.
753:
754: With the increase of coupling parameter the range of synchronized
755: time scales $[s_1, s_2]$, for which the TSS conditions
756: (\ref{eq:PhaseLocking}) and (\ref {eq:SynchroEnergy}) are
757: satisfied, is expanded. Time scales which are close to the most
758: intensive scales in wavelet power spectrum and whose energy is
759: large enough, start to be involved in synchronous dynamics.
760:
761: The latter is well illustrated by Fig.\,\ref{fgr:ScaleDifference}
762: from which it can be seen, that for both base time scales $s_b$
763: and $s_m$ extension of the range of phase locked scales $\Delta s$
764: with the increase of $R$ is observed.
765: %
766: At large values of coupling coefficient $R$ the range of
767: synchronous scales $\Delta s_b=s_{2b}-s_{1b}$ in the area of the
768: base scale $s_b$ ceases to be expanded and, moreover, is narrowed
769: down at $R>0.5$.
770: %
771: At the same time in the area of scales $s_m$, describing
772: modulation of amplitude of the output field of BWO, the range of
773: synchronous scales $\Delta s_m=s_{2m}-s_{1m}$ increases linearly
774: with the growth of $R$, at the expense of growth of top bound of
775: area of synchronized scales $s_{2m}$.
776: %
777: At last we must note that with growth of $R$, the energy of the
778: wavelet power spectrum of a response system is redistributed in
779: such a manner that the increase of energy of larger scales $s>s_m
780: $ is observed.
781: %
782: Then the most intensive time scales in wavelet spectrum appear
783: locally, near the top bound $s_{2m}$ of synchronized time scales,
784: therefore these time scales are first to be synchronized with the
785: increase of coupling coefficient $R$.
786:
787:
788:
789: \section{Measure of synchronization of spatially extended electron--wave system}
790: \label{Sct:Measure}
791:
792: Introducing continuous set of time scales $s$ and the
793: instantaneous phases associated with them, as well as separation
794: of the range of synchronous time scales $\Delta s=s_2-s_1$ allows
795: us to inject the quantitative performance of a measure of chaotic
796: synchronization of coupled systems
797: \cite{Koronovskii:2004_JETPLettersEngl, Hramov:2004_Chaos}. This
798: measure $\gamma$ can be defined as a part of wavelet spectrum
799: energy being fallen on the synchronized time scales
800: \begin{equation}
801: \gamma=\frac{1}{E_{2}}\int\limits_{s_m}^{s_b}\langle
802: E_{2}(s)\rangle\,ds, \label{eq:Gamma}
803: \end{equation}
804: where $[s_m;s_b]$ is the range of time scales for which the
805: condition~(\ref{eq:PhaseLocking}) is fulfilled and $E_{2}$ is the
806: full energy of the wavelet spectrum
807: \begin{equation}
808: E_{2}=\int\limits_{0}^{+\infty}\langle E_{2}(s)\rangle\,ds
809: \end{equation}
810: of the response system.
811:
812: The energy distribution over different time scales is a very
813: useful characteristics for complex system behavior description. It
814: is widely used in different applications (see e.g.,
815: \cite{Schwarz:1998_WaveletsAstro, WaveletsInPhysics:2004}).
816:
817: This measure $\gamma$ is equal to zero for the nonsynchronized
818: oscillations. $\gamma\not=0$ means that in the coupled systems the
819: conditions of TSS (\ref{eq:PhaseLocking}) and
820: (\ref{eq:SynchroEnergy}) are implemented. The value $\gamma=1$
821: shows, that oscillations in each of subsystems are identical or
822: shifted upon some lag of time $T_\tau$. This is a complete or lag
823: chaotic synchronization \cite{Pikovsky:2002_SynhroBook}. Increase
824: of $\gamma$ from $0$ up to $1$ testifies to the increase of a part
825: of the wavelet spectrum energy, being fallen on synchronous time
826: scales $s$. Actually, the measure $\gamma$ characterizes how close
827: to each other the chaotic oscillations are in both spatially
828: extended systems.
829:
830: In our case when the synchronous behaviour is observed for two
831: base scales $s_m$ and $s_b$, it is possible to introduce measures
832: of synchronization $\gamma_b$ and $\gamma_m$ according to each of
833: time scales $s_b$ and $s_m$, and also the integral performance of
834: the TSS measure of the systems: $\gamma=\gamma_{b} + \gamma_{m}$.
835: Dependence of the measure $\gamma$ on the coupling coefficient $R$
836: characterizes degree of chaotic synchronization of oscillations in
837: the electron--wave systems with a backward wave.
838:
839: \begin{figure}
840: \centerline{\scalebox{0.4}{\includegraphics{Fig5.ps}}}
841:
842: \caption{The dependence of the synchronization measures
843: $\gamma_b$, $\gamma_m$ and $\gamma$ of the response system on the
844: coupling strength $R$
845: \label{fgr:Gamma}}
846: \end{figure}
847:
848:
849: In Fig.\,\ref{fgr:Gamma} the dependence of the synchronization
850: measures $\gamma_b$, $\gamma_m$ and $\gamma$ for the response
851: systems upon the coupling parameter $R$ is shown. It is obvious,
852: that at small $R<0.4\div0.5 $ the main part of energy of the
853: wavelet power spectrum, being fallen at synchronized time scales,
854: corresponds to the base scale $s_b$ ($\gamma_b>\gamma_m$).
855:
856: Thus, in the coupled systems there is the capture of phases of
857: time scales $s\sim s_b$ close to the time scale corresponding to
858: the frequency of synchronism of electronic and electromagnetic
859: waves. With further growth of $R$ the capture of frequencies close
860: to the modulating frequency $f_m$ of the output signal of the
861: driving BWO takes place. With the increase of $R$ the range of the
862: synchronized scales describing low-frequency chaotic modulation of
863: the output signal also increases. Hence, $\gamma_m$ (the
864: normalized part of energy of synchronized scales corresponding to
865: low-frequency modulation oscillations) grows with the increase of
866: $R$. At $R> 0.5$ part of energy of synchronous low-frequency
867: modulation time scales $\gamma_m$ becomes more than the value
868: $\gamma_b$, describing part of energy of synchronous
869: high-frequency oscillations.
870:
871:
872: The integral part of energy $\gamma$, being fallen on synchronous
873: time scales, monotonically increases with the growth of value of
874: coupling coefficient $R$ [see Fig.\,\ref {fgr:Gamma}]. However,
875: even at large values of $R$, when the response active medium is
876: hardly influenced by the signal of the driving system
877: ($R^2\sim0.3\div0.4 $), the part of the energy, being fallen upon
878: synchronous time scales, does not exceed value $\gamma\sim0.3 $.
879:
880: Thus, at chaotic synchronization of two unidirectional coupled BWO
881: at the increase of coupling coefficient, we can observe TSS near
882: the most intensive time scale $s_b$ corresponding to frequency
883: $f_b$ of synchronism of electronic and electromagnetic waves. With
884: the growth of $R$ there is an increase of energy of synchronized
885: scales in the area of the base scale $s_b$. At the same time, a
886: range of synchronized time scales $\Delta s_m$, corresponding to
887: low-frequency modulation of oscillations of field $F$, appears.
888:
889: With the further increase of $R$ the growth of wavelet power
890: spectrum energy, being fallen on synchronous time scales, is due
891: only to the increase of the range of synchronized time scales
892: $\Delta s_m$. The given behaviour of two unidirectional coupled
893: systems is quantitatively described by the measure of chaotic
894: synchronization $\gamma=\gamma(R)$.
895:
896:
897: \section{Space--time dynamics of response chaotic electron--wave system}
898: \label{Sct:SpaceDynamics}
899:
900: Let us consider the physical processes in the non--autonomous
901: (response) active medium when the chaotic synchronization appears.
902: First of all, we are interested in space dynamics of response
903: active medium.
904:
905: The external chaotic signal generated by the drive distributed
906: system influences on the point with coordinate $\xi=L$ of the
907: response active medium and then spreads towards electron wave to
908: the output $\xi=0$ border. The dimensionless length of the active
909: media denoted as $L$ is equal to $1$. The synchronization of the
910: whole active electron--wave system discussed above means the time
911: scale synchronization of external signal $F_{ext1}({\tau})$ and
912: output signal $F_{out2}(\tau)=F_2(\tau,\xi=0)$ of response media
913: takes place. To consider the space--time dynamics of response
914: system we have analyzed the presence of time scale synchronization
915: between the external signal and the time series $F_2(\tau,\xi)$
916: obtained from different points $0<\xi<L$ of interaction space.
917:
918: Let us consider the synchronization on time scales which are close
919: to the most intensive scale $s_b$ of external chaotic signal
920: wavelet spectrum. The synchronization of these time scales appears
921: when the value of coupling strength $R$ is small enough and when
922: the energy of external signal $P_{ext}=R^2|F_{ext1}|^2$ is also
923: small (see Section~\ref{Sct:Synchro} for detail). The range
924: $\Delta s_b$ of synchronized time scales may be considered as a
925: function of interaction space coordinate of active medium. This
926: function ${\Delta s_b=\Delta s_b(\xi)}$ as well as the
927: synchronization measure ${\gamma=\gamma(\xi)}$ (\ref{eq:Gamma})
928: are the characteristics describing the space dynamics of chaotic
929: synchronization appearance when the external chaotic signal acts
930: on one of the border of the distributed system.
931:
932: Fig.~\ref{fgr:DsG_xi}({a}) shows areas of synchronized time scales
933: on the plane ``coordinate $(L-\xi)$ --- time scales $s$'' for
934: three different values of coupling strength $R$. In
935: Fig.~\ref{fgr:DsG_xi}(a) one can see that the width of the
936: synchronized time scales area is maximum in the points of
937: interaction space which are close to the border of system on which
938: the chaotic signal of the drive active medium acts (see the border
939: conditions (\ref{ref_bwopp_boundary_properties_TWO2}) for the
940: field $F_2$ of the response system). When the coordinate $\xi$
941: decreases, the area of synchronized time scales $\Delta s_b(\xi)$
942: converges gradually. The measure of synchronization
943: $\gamma_b(\xi)$ behaves along the interaction space in the similar
944: way [see Fig.~\ref{fgr:DsG_xi}(b)]. With the decreasing of
945: $\xi$--coordinate the part of wavelet spectra energy being fallen
946: on the synchronized time scales $\Delta s_b(\xi)$ also decreases.
947:
948: \begin{figure}
949: \centerline{\scalebox{0.5}{\includegraphics{Fig6.ps}}}
950: \caption{(a) The areas of synchronized time scales on the plane
951: ``coordinate $(L-\xi)$ --- time scales $s$''. The coordinate
952: ${(L-\xi)}$ is counted from the end of the response system. (b)
953: The dependence of synchronization measure $\gamma_b$ on the
954: coordinate of interaction space $(L-\xi)$. The black squares
955: ($\blacksquare$) correspond to the value of coupling parameter
956: $R=0.4$, circles ($\circ$) --- $R=0.3$ and black triangles
957: ($\blacktriangle$) --- $R=0.2$ \label{fgr:DsG_xi}}
958: \end{figure}
959:
960:
961: When the coupling strength $R$ is rather small ($R<0.29$) and the
962: synchronization of two coupled active media is not observed [see
963: Sections~\ref{Sct:Synchro} and \ref{Sct:Measure}], the width of
964: the synchronized time scales area becomes equal to zero at the
965: point of interaction space $\xi_s<L$. It means that the intricate
966: space dynamics of phase locking takes place in the asynchronous
967: regime. The dependence of the width of the synchronized time
968: scales area on the coordinate $\xi$ (see also
969: Fig.~\ref{fgr:DsG_xi}(a), case $R=0.2$) is described by
970: \begin{equation}
971: \begin{array}{ll}
972: \Delta s\not=0 &
973: \text{~ïðè~}\xi>\xi_s, \\
974: \Delta s=0 & \text{~ïðè~}\xi<\xi_s. \\
975: \end{array}
976: \label{eq:Ds_xi}
977: \end{equation}
978: So, in this case it is possible to say that the interaction space
979: of the response active medium is separated to two regions.
980:
981: First of them is characterized by length $L_s=L-\xi_s$ and abuts
982: on the border $\xi=L$ of the response active system. The chaotic
983: oscillations $F_2(\tau,\xi)$ ($\xi\in(\xi_s,L)$) in this region
984: are synchronous (in terms of time scale synchronization) with the
985: external chaotic signal $F_{ext1}(\tau)$ generated by the drive
986: system. Therefore, the synchronization measure is not equal to
987: zero: $\left.\gamma_b\right|_{\xi>\xi_s}\not=0$. This region of
988: interaction space has naturally been called as ``the region of
989: synchronous oscillation''. Its length $L_s$ has been called
990: ``length of synchronization''.
991:
992: In the second region $\xi\in(0,\xi_s)$ the chaotic synchronization
993: is destroyed and the range of the synchronized time
994: scales~(\ref{eq:Ds_xi}) as well as the synchronization measure is
995: equal to zero.
996:
997: With the increasing of $R$ the synchronization length also
998: increases. For some value $R=R_b$ the length of synchronization
999: becomes equal to the length of the interaction space $\L_s\equiv
1000: L$ (see Section~\ref{Sct:Synchro} and
1001: Fig.~\ref{fgr:ScaleDifference}). This means that in all space of
1002: response of electron--wave medium the chaotic synchronization
1003: takes place and, therefore, one can say about synchronization of
1004: considered systems as whole. This regime of chaotic
1005: synchronization has been described above in
1006: sections~\ref{Sct:Synchro} and \ref{Sct:Measure} when the time
1007: scale dynamics of chaotic signal get from output $\xi=0$ of
1008: response system has been considered.
1009:
1010:
1011: Fig.~\ref{fgr:DsG_xi}(a) corresponding to the case of
1012: synchronization of the whole response active medium shows the
1013: borders of synchronized time scales $s$ along the interaction
1014: space. The value of coupling strength has been selected as
1015: $R=0.3\approx R_b$. There is the range $\Delta s$ of synchronized
1016: time scales for every point $0\leq\xi\leq L$. With coupling
1017: strength increasing the range of synchronized time scales also
1018: increases [see Fig.~\ref{fgr:DsG_xi}(a), $R=0.4$]. This result
1019: agrees well with Sec.~\ref{Sct:Synchro}. The value of wavelet
1020: power spectrum energy being fallen on synchronized time scales is
1021: equal approximately to $30\div40\%$ in the point with coordinate
1022: $\xi=0.99L$ and decreases to $1\div5\%$ at the output border
1023: $\xi=0$ of the response active media.
1024:
1025:
1026: In Fig.~\ref{fgr:Xis_R} the dependence of the length of
1027: synchronization $L_s$ on coupling parameter $R$ for the base time
1028: scale $s_b$ is shown (solid line). It is clear, that with the
1029: increasing of coupling strength the length of synchronization
1030: grows monotonically according to the linear law: $L_s(R)\sim R$.
1031: It follows, that with the growth of coupling coefficient $R$ we
1032: have the linear increase of the region where time scale
1033: synchronization near base scale $s_b$ is observed. Such behaviour
1034: of the response system is determined by the features of
1035: interaction of electronic and electromagnetic waves. Destruction
1036: of time scale synchronization is defined by violation of phase
1037: ratio between backward waves of electron current $I(\tau,\xi)$
1038: and electromagnetic field $F(\tau,\xi)$.
1039:
1040: \begin{figure}
1041: \centerline{\scalebox{0.3}{\includegraphics{Fig7.ps}}}
1042: \caption{Length of synchronization of response spatially extended
1043: system as function of coupling strength $R$. The dependence for
1044: base time scale $s_b$ is pointed by circles and approximation
1045: $L_s\sim R$ is plotted by solid line. The numerical dependence
1046: and approximation for modulating time scale $s_m$ is pointed by
1047: squares and dashed line \label{fgr:Xis_R}}
1048: \end{figure}
1049:
1050:
1051: So, at the fixed length of the response system, the phase shift
1052: $\Delta\varphi$ between field and current waves, originating due
1053: to the influence of external signal upon the response system,
1054: reduces on the length of synchronization $L_s$ to destruction of
1055: time scale synchronization on some time scale $s$. The phase shift
1056: is defined as $\Delta\varphi=2\pi|f_b-f_s|\xi_s/v_g$, where
1057: $f_b=1/s_b $, $f_s=1/s$ and $v_g$ is the the group velocity of the
1058: backward electromagnetic wave.
1059:
1060: On the other hand, dependence of the phase shift $\Delta\varphi$
1061: upon the amplitude of the external field $R|F_2|$ can be presented
1062: as $\Delta\varphi=\kappa RF_2$, where $\kappa$ is the coefficient
1063: of proportionality depending on length of the response spatially
1064: extended system $A_2$. Comparing the estimated expressions for the
1065: phase shift $\Delta\varphi$ it is found out that the length of
1066: synchronization $L_s$ is proportional to $R$, as the numerical
1067: simulation shows (see Fig.~\ref{fgr:Xis_R}).
1068:
1069: It is important to note that curves shown in Fig.~\ref{fgr:DsG_xi}
1070: correspond to the base time scale $s_b$ and time scales close to
1071: it. The similar results have been obtained for the space--time
1072: dynamics of scales which are close to the modulation time scale
1073: $s_m$ [see Fig.~\ref{fgr:Xis_R}, dashed line]. The difference
1074: consists in that in this case the length of synchronized time
1075: scales area (length of synchronization) becomes equal to the
1076: system length for the bigger values of coupling strength than for
1077: the base time scale $s_b$. This also agrees well with the results
1078: discussed above.
1079:
1080:
1081:
1082: \section{Conclusion}
1083: \label{Sct:Conclusion}
1084:
1085: In conclusion, in our article we have considered the chaotic
1086: oscillations in two unidirectionally coupled electron--wave active
1087: media. The approach of chaotic synchronization analysis proposed
1088: in our work~\cite{Koronovskii:2004_JETPLettersEngl,
1089: Hramov:2004_Chaos} has been used. This approach is based on the
1090: consideration of continuous set of time scales and chaotic signal
1091: phases associated with them. The application of this method allows
1092: us to investigate the chaotic synchronization of distributed
1093: active media demonstrating the intricate dynamics whereas the
1094: traditional methods of chaotic signal phase introduction fail due
1095: to the complex topology of the attractor. At the same time we
1096: succeed in dividing the chaotic synchronization into
1097: synchronization of low--frequency modulation oscillations of
1098: electromagnetic field and high--frequency generation one. One can
1099: see that it was impossible to do it using traditional approaches
1100: of chaotic synchronization analysis. So, the results described in
1101: our work allow to understand more clearly the common mechanisms of
1102: chaotic synchronization and justify the efficiency of new
1103: TSS-approach to chaotic synchronization analysis.
1104:
1105: The main results are the following. With the increase of coupling
1106: strength the chaotic synchronization regime appears and is
1107: characterized by phase relationship between time scales of chaotic
1108: oscillations of drive and response active media. First of all, the
1109: synchronization of time scales corresponding to high--frequency
1110: spectral component appears. The synchronization of time scales
1111: corresponding to low--frequency oscillations takes place with the
1112: larger values of coupling strength $R$. When the chaotic
1113: synchronization of the whole active medium is not observed the
1114: interaction space may be divided into two regions corresponding to
1115: synchronous and asynchronous oscillations, respectively. The
1116: length of synchronous oscillation area $L_s$ grows when the
1117: coupling strength $R$ increases. The appearance of chaotic
1118: synchronization of two unidirectionally coupled active media is
1119: determined by condition of equality of synchronization length and
1120: length of the system.
1121:
1122: It should also be noted that the method of investigation of
1123: chaotic synchronization of two coupled distributed systems
1124: discussed above can be easily applied to the analysis of higher
1125: spatial dimensions systems, such as 2D and 3D ones. In this case
1126: one have to analyze time series $x(t,\mathbf{r})$ obtained from
1127: different points $\mathbf{r}$ of space in the manner discussed
1128: above.
1129:
1130:
1131:
1132: \section{Acknowledgments}
1133: \label{Sct:Acknowledgments} We express our appreciation to
1134: Professor Dmirtiy~I.~Trubetskov for valuable discussions. We also
1135: thank Dr. S.V.~Eremina for language support.
1136: %
1137: This work has been supported by U.S.~Civilian Research \&
1138: Development Foundation for the Independent States of the Former
1139: Soviet Union (CRDF, grant {REC--006}) and Russian Foundation for
1140: Basic Research (grant 05--02--16273). We also thank ``Dynasty''
1141: Foundation.
1142:
1143:
1144:
1145:
1146: %\bibliography{Aeh}
1147:
1148: \begin{thebibliography}{46}
1149: \expandafter\ifx\csname
1150: natexlab\endcsname\relax\def\natexlab#1{#1}\fi
1151: \expandafter\ifx\csname bibnamefont\endcsname\relax
1152: \def\bibnamefont#1{#1}\fi
1153: \expandafter\ifx\csname bibfnamefont\endcsname\relax
1154: \def\bibfnamefont#1{#1}\fi
1155: \expandafter\ifx\csname citenamefont\endcsname\relax
1156: \def\citenamefont#1{#1}\fi
1157: \expandafter\ifx\csname url\endcsname\relax
1158: \def\url#1{\texttt{#1}}\fi
1159: \expandafter\ifx\csname
1160: urlprefix\endcsname\relax\def\urlprefix{URL }\fi
1161: \providecommand{\bibinfo}[2]{#2}
1162: \providecommand{\eprint}[2][]{\url{#2}}
1163:
1164: \bibitem[{\citenamefont{{Rosenblum~M.G., Pikovsky~A.S.,
1165: Kurths~J.}}(1996)}]{Rosenblum:1996_PhaseSynchro}
1166: \bibinfo{author}{\bibnamefont{{Rosenblum~M.G., Pikovsky~A.S., Kurths~J.}}},
1167: \bibinfo{journal}{Phys. Rev. Lett.} \textbf{\bibinfo{volume}{76}},
1168: \bibinfo{pages}{1804} (\bibinfo{year}{1996}).
1169:
1170: \bibitem[{\citenamefont{{Pikovsky~A., Rosenblum~M.,
1171: Kurths~J.}}(2000)}]{Pikovsky:2000_SynchroReview}
1172: \bibinfo{author}{\bibnamefont{{Pikovsky~A., Rosenblum~M., Kurths~J.}}},
1173: \bibinfo{journal}{Int. J. Bifurcation and Chaos}
1174: \textbf{\bibinfo{volume}{10}}, \bibinfo{pages}{2291} (\bibinfo{year}{2000}).
1175:
1176: \bibitem[{\citenamefont{{Boccaletti S., Kurths J., Osipov G., Valladares D.L.,
1177: Zhou C.}}(2002)}]{Boccaletti:2002_ChaosSynchro}
1178: \bibinfo{author}{\bibnamefont{{Boccaletti S., Kurths J., Osipov G., Valladares
1179: D.L., Zhou C.}}}, \bibinfo{journal}{Physics Reports}
1180: \textbf{\bibinfo{volume}{366}}, \bibinfo{pages}{1} (\bibinfo{year}{2002}).
1181:
1182: \bibitem[{\citenamefont{{Pikovsky~A., Rosenblum~M.,
1183: Kurths~J.}}(2001)}]{Pikovsky:2002_SynhroBook}
1184: \bibinfo{author}{\bibnamefont{{Pikovsky~A., Rosenblum~M., Kurths~J.}}},
1185: \emph{\bibinfo{title}{Synhronization: a universal concept in nonlinear
1186: sciences}} (\bibinfo{publisher}{Cambridge University Press},
1187: \bibinfo{year}{2001}).
1188:
1189: \bibitem[{\citenamefont{{Ticos~C.M., Rosa~E., Pardo~W.B., Walkenstein~J.A.,
1190: Monti~M.}}(2000)}]{Ticos:2000_PlasmaDischarge}
1191: \bibinfo{author}{\bibnamefont{{Ticos~C.M., Rosa~E., Pardo~W.B.,
1192: Walkenstein~J.A., Monti~M.}}}, \bibinfo{journal}{Phys. Rev. Lett.}
1193: \textbf{\bibinfo{volume}{85}}, \bibinfo{pages}{2929} (\bibinfo{year}{2000}).
1194:
1195: \bibitem[{\citenamefont{{Anishchenko~V.S., Balanov~A.G., Janson~N.B.,
1196: Igosheva~N.B., Bordyugov~G.V.}}(2000)}]{Anishchenko:2000_humanSynchro}
1197: \bibinfo{author}{\bibnamefont{{Anishchenko~V.S., Balanov~A.G., Janson~N.B.,
1198: Igosheva~N.B., Bordyugov~G.V.}}}, \bibinfo{journal}{Int. J. Bifurcation and
1199: Chaos} \textbf{\bibinfo{volume}{10}}, \bibinfo{pages}{2339}
1200: (\bibinfo{year}{2000}).
1201:
1202: \bibitem[{\citenamefont{{Tass~P.A. et al.}}(2003)}]{Tass:2003_NeuroSynchro}
1203: \bibinfo{author}{\bibnamefont{{Tass~P.A. et al.}}}, \bibinfo{journal}{Phys.
1204: Rev. Lett.} \textbf{\bibinfo{volume}{90}}, \bibinfo{pages}{088101}
1205: (\bibinfo{year}{2003}).
1206:
1207: \bibitem[{\citenamefont{{Prokhorov M. at
1208: al.}}(2003)}]{Prokhorov:2003_HumanSynchroPRE}
1209: \bibinfo{author}{\bibnamefont{{Prokhorov M. at al.}}}, \bibinfo{journal}{Phys.
1210: Rev. E} \textbf{\bibinfo{volume}{68}}, \bibinfo{pages}{041913}
1211: (\bibinfo{year}{2003}).
1212:
1213: \bibitem[{\citenamefont{{Murali~K.,
1214: Lakshmanan~M.}}(1994)}]{Murali:1993_SignalTransmission}
1215: \bibinfo{author}{\bibnamefont{{Murali~K., Lakshmanan~M.}}},
1216: \bibinfo{journal}{Phys. Rev. E} \textbf{\bibinfo{volume}{48}},
1217: \bibinfo{pages}{R1624} (\bibinfo{year}{1994}).
1218:
1219: \bibitem[{\citenamefont{{Yang~T., Wu~C.W. and
1220: Chua~L.O.}}(1997)}]{Chua:1997_Criptography}
1221: \bibinfo{author}{\bibnamefont{{Yang~T., Wu~C.W. and Chua~L.O.}}},
1222: \bibinfo{journal}{IEEE Trans. Circuits and Syst.}
1223: \textbf{\bibinfo{volume}{44}}, \bibinfo{pages}{469} (\bibinfo{year}{1997}).
1224:
1225: \bibitem[{\citenamefont{{Liu Yaowen, Ge Guangming, Zhao Hong, Wang
1226: Yinghai}}(2000)}]{Yaowen:2002_ChaoticCommunication}
1227: \bibinfo{author}{\bibnamefont{{Liu Yaowen, Ge Guangming, Zhao Hong, Wang
1228: Yinghai}}}, \bibinfo{journal}{Phys. Rev. E} \textbf{\bibinfo{volume}{62}},
1229: \bibinfo{pages}{7898} (\bibinfo{year}{2000}).
1230:
1231: \bibitem[{\citenamefont{{Garcia-Ojalvo J., Roy
1232: R.}}(2001)}]{Garcia-Ojalvo:2001_ChaosCommunication}
1233: \bibinfo{author}{\bibnamefont{{Garcia-Ojalvo J., Roy R.}}},
1234: \bibinfo{journal}{Phys.Rev.Lett.} \textbf{\bibinfo{volume}{86}},
1235: \bibinfo{pages}{5204} (\bibinfo{year}{2001}).
1236:
1237: \bibitem[{\citenamefont{{Kurths J., Pikovsky
1238: A.S.}}(1995)}]{Kurths:1995_DistrSystems}
1239: \bibinfo{author}{\bibnamefont{{Kurths J., Pikovsky A.S.}}},
1240: \bibinfo{journal}{Chaos Solit. Fract.} \textbf{\bibinfo{volume}{5}},
1241: \bibinfo{pages}{1893} (\bibinfo{year}{1995}).
1242:
1243: \bibitem[{\citenamefont{{Boccaletti S., Bragard J., Arecchi
1244: F.T.}}(1999)}]{Boccaletti:1999_ControllingChaos}
1245: \bibinfo{author}{\bibnamefont{{Boccaletti S., Bragard J., Arecchi F.T.}}},
1246: \bibinfo{journal}{Phys. Rev. E.} \textbf{\bibinfo{volume}{59}},
1247: \bibinfo{pages}{6574} (\bibinfo{year}{1999}).
1248:
1249: \bibitem[{\citenamefont{{Bragard J., Arecchi F.T., Boccaletti
1250: S.}}(2000)}]{Bragard:2000}
1251: \bibinfo{author}{\bibnamefont{{Bragard J., Arecchi F.T., Boccaletti S.}}},
1252: \bibinfo{journal}{Int. J. Bifurcation and Chaos}
1253: \textbf{\bibinfo{volume}{10}}, \bibinfo{pages}{2381} (\bibinfo{year}{2000}).
1254:
1255: \bibitem[{\citenamefont{{Ticos C.M., Rosa E., Pardo W.B., Walkenstein J.A.,
1256: Monti M.}}(2000)}]{Ticos:2000_Plasma}
1257: \bibinfo{author}{\bibnamefont{{Ticos C.M., Rosa E., Pardo W.B., Walkenstein
1258: J.A., Monti M.}}}, \bibinfo{journal}{Phys.Rev.Lett.}
1259: \textbf{\bibinfo{volume}{85}} (\bibinfo{year}{2000}).
1260:
1261: \bibitem[{\citenamefont{{Ahlers V., Pikovsky A.}}(2002)}]{Ahlers:2002}
1262: \bibinfo{author}{\bibnamefont{{Ahlers V., Pikovsky A.}}},
1263: \bibinfo{journal}{Phys.Rev.Lett.} \textbf{\bibinfo{volume}{88}},
1264: \bibinfo{pages}{254101} (\bibinfo{year}{2002}).
1265:
1266: \bibitem[{\citenamefont{{Valladares D.L., Boccaletti S., Feudel F., Kurths
1267: J.}}(2002)}]{Valladares:2002}
1268: \bibinfo{author}{\bibnamefont{{Valladares D.L., Boccaletti S., Feudel F.,
1269: Kurths J.}}}, \bibinfo{journal}{Phys. Rev. E} \textbf{\bibinfo{volume}{65}},
1270: \bibinfo{pages}{055208} (\bibinfo{year}{2002}).
1271:
1272: \bibitem[{\citenamefont{{Mendoza C., Boccaletti S., Politi
1273: A.}}(2004)}]{Mendoza:2004}
1274: \bibinfo{author}{\bibnamefont{{Mendoza C., Boccaletti S., Politi A.}}},
1275: \bibinfo{journal}{Phys. Rev. E.} \textbf{\bibinfo{volume}{69}},
1276: \bibinfo{pages}{047202} (\bibinfo{year}{2004}).
1277:
1278: \bibitem[{\citenamefont{{Pecora L.M., Carroll T.L., Jonson G.A., Mar
1279: D.J.}}(1997)}]{Pecora:1997_SynchroChaos}
1280: \bibinfo{author}{\bibnamefont{{Pecora L.M., Carroll T.L., Jonson G.A., Mar
1281: D.J.}}}, \bibinfo{journal}{Chaos} \textbf{\bibinfo{volume}{7}},
1282: \bibinfo{pages}{520} (\bibinfo{year}{1997}).
1283:
1284: \bibitem[{\citenamefont{{Boccaletti~S.,
1285: Valladares~D.L.}}(2000)}]{Boccaletti:2000_IntermitLagSynchro}
1286: \bibinfo{author}{\bibnamefont{{Boccaletti~S., Valladares~D.L.}}},
1287: \bibinfo{journal}{Phys. Rev. E} \textbf{\bibinfo{volume}{62}},
1288: \bibinfo{pages}{7497} (\bibinfo{year}{2000}).
1289:
1290: \bibitem[{\citenamefont{{Anishchenko~V.S.,
1291: Vadivasova~T.E.}}(2002)}]{Anishchenko:2002_SynchroEng}
1292: \bibinfo{author}{\bibnamefont{{Anishchenko~V.S., Vadivasova~T.E.}}},
1293: \bibinfo{journal}{Journal of Communications Technology and Electronics}
1294: \textbf{\bibinfo{volume}{47}}, \bibinfo{pages}{117} (\bibinfo{year}{2002}).
1295:
1296: \bibitem[{\citenamefont{{Anishchenko~V.S.,
1297: Vadivasova~T.E.}}(2004)}]{Anishchenko:2004_ChaosSynchro}
1298: \bibinfo{author}{\bibnamefont{{Anishchenko~V.S., Vadivasova~T.E.}}},
1299: \bibinfo{journal}{Journal of Communications Technology and Electronics}
1300: \textbf{\bibinfo{volume}{49}} (\bibinfo{year}{2004}).
1301:
1302: \bibitem[{\citenamefont{{Hramov A.E., Koronovskii
1303: A.A.}}(2004)}]{Hramov:2004_Chaos}
1304: \bibinfo{author}{\bibnamefont{{Hramov A.E., Koronovskii A.A.}}},
1305: \bibinfo{journal}{Chaos} \textbf{\bibinfo{volume}{14}}, \bibinfo{pages}{603}
1306: (\bibinfo{year}{2004}).
1307:
1308: \bibitem[{\citenamefont{{Osipov G.V., Hu B., Zhou C., Ivanchenko M.V., Kurths
1309: J.}}(2003)}]{Osipov:2003_Synchro_Chaos}
1310: \bibinfo{author}{\bibnamefont{{Osipov G.V., Hu B., Zhou C., Ivanchenko M.V.,
1311: Kurths J.}}}, \bibinfo{journal}{Phys.Rev.Lett.}
1312: \textbf{\bibinfo{volume}{91}}, \bibinfo{pages}{024101}
1313: (\bibinfo{year}{2003}).
1314:
1315: \bibitem[{\citenamefont{{Romano M.C., Thiel M., Kurths J., von Bloh
1316: W.}}(2004)}]{Romano:2004_Phys.Lett.A}
1317: \bibinfo{author}{\bibnamefont{{Romano M.C., Thiel M., Kurths J., von Bloh
1318: W.}}}, \bibinfo{journal}{Phys. Lett. A} \textbf{\bibinfo{volume}{330}},
1319: \bibinfo{pages}{214} (\bibinfo{year}{2004}).
1320:
1321: \bibitem[{\citenamefont{{Koronovskii A.A., Hramov
1322: A.E.}}(2004)}]{Koronovskii:2004_JETPLettersEngl}
1323: \bibinfo{author}{\bibnamefont{{Koronovskii A.A., Hramov A.E.}}},
1324: \bibinfo{journal}{JETP Lett.} \textbf{\bibinfo{volume}{79}},
1325: \bibinfo{pages}{316} (\bibinfo{year}{2004}).
1326:
1327: \bibitem[{\citenamefont{{Daubechies~I.}}(1992)}]{Daubechies:1992_WVTBook}
1328: \bibinfo{author}{\bibnamefont{{Daubechies~I.}}}, \emph{\bibinfo{title}{Ten
1329: lectures on wavelets}} (\bibinfo{publisher}{SIAM}, \bibinfo{year}{1992}).
1330:
1331: \bibitem[{\citenamefont{{Kaiser~G.}}(1994)}]{Kaiser:1994_Wvt}
1332: \bibinfo{author}{\bibnamefont{{Kaiser~G.}}}, \emph{\bibinfo{title}{A friendly
1333: guide to wavelets}} (\bibinfo{publisher}{Springer Verlag},
1334: \bibinfo{year}{1994}).
1335:
1336: \bibitem[{\citenamefont{{Koronovskii~A.A.,
1337: Hramov~A.E.}}(2003)}]{alkor:2003_WVTBookEng}
1338: \bibinfo{author}{\bibnamefont{{Koronovskii~A.A., Hramov~A.E.}}},
1339: \emph{\bibinfo{title}{Continuous wavelet analysis and its applications (In
1340: Russian)}} (\bibinfo{publisher}{Moscow, Fizmatlit}, \bibinfo{year}{2003}).
1341:
1342: \bibitem[{\citenamefont{{Grossman~A. and
1343: Morlet~J.}}(1984)}]{Grossman:1984_Morlet}
1344: \bibinfo{author}{\bibnamefont{{Grossman~A. and Morlet~J.}}},
1345: \bibinfo{journal}{SIAM J. Math. Anal.} \textbf{\bibinfo{volume}{15}},
1346: \bibinfo{pages}{273} (\bibinfo{year}{1984}).
1347:
1348: \bibitem[{\citenamefont{{Soohoo
1349: R.F.}}(1971)}]{Soohoo:1971_MicrowaveElectronicsBook}
1350: \bibinfo{author}{\bibnamefont{{Soohoo R.F.}}}, \emph{\bibinfo{title}{Microwave
1351: Electronics}} (\bibinfo{publisher}{Addison-Wesley Longman},
1352: \bibinfo{year}{1971}).
1353:
1354: \bibitem[{boo(1983)}]{book:1983_microwave}
1355: \emph{\bibinfo{title}{Infrared and Millimeter Waves}}
1356: (\bibinfo{publisher}{Academic Press, New York}, \bibinfo{year}{1983}),
1357: \bibinfo{edition}{{ K. Button}} ed.
1358:
1359: \bibitem[{\citenamefont{{Granatstein V.L., Alexeff
1360: I.}}(1987)}]{Granatstein:1987_Book}
1361: \bibinfo{author}{\bibnamefont{{Granatstein V.L., Alexeff I.}}},
1362: \emph{\bibinfo{title}{High Power Microwave Sources}}
1363: (\bibinfo{publisher}{Artech House Microwave Library}, \bibinfo{year}{1987}).
1364:
1365: \bibitem[{\citenamefont{{Felch K.L., Danly B.G., Jory H.R. et
1366: al}}(1999)}]{Felch:1999_SmallAuthor}
1367: \bibinfo{author}{\bibnamefont{{Felch K.L., Danly B.G., Jory H.R. et al}}},
1368: \bibinfo{journal}{Proceedings IEEE} \textbf{\bibinfo{volume}{87}},
1369: \bibinfo{pages}{752} (\bibinfo{year}{1999}).
1370:
1371: \bibitem[{\citenamefont{{Bezruchko B.P., Kuznetsov S.P., Trubetskov
1372: D.I.}}(1979)}]{Bezruchko:1979_BWOChaos}
1373: \bibinfo{author}{\bibnamefont{{Bezruchko B.P., Kuznetsov S.P., Trubetskov
1374: D.I.}}}, \bibinfo{journal}{JETP Lett.} \textbf{\bibinfo{volume}{29}},
1375: \bibinfo{pages}{162} (\bibinfo{year}{1979}).
1376:
1377: \bibitem[{\citenamefont{{Ginzburg N.S., Kuznetsov S.P., Fedoseeva
1378: T.N.}}(1979)}]{Ginzburg:1979_engl}
1379: \bibinfo{author}{\bibnamefont{{Ginzburg N.S., Kuznetsov S.P., Fedoseeva
1380: T.N.}}}, \bibinfo{journal}{Radiophys. Quantum Electron}
1381: \textbf{\bibinfo{volume}{21}}, \bibinfo{pages}{728} (\bibinfo{year}{1979}).
1382:
1383: \bibitem[{\citenamefont{{Levush B., Antonsen T.M., Bromborsky A., Lou W.R.,
1384: Carmel Y.}}(1992)}]{Levush:1992_BWO}
1385: \bibinfo{author}{\bibnamefont{{Levush B., Antonsen T.M., Bromborsky A., Lou
1386: W.R., Carmel Y.}}}, \bibinfo{journal}{IEEE Trans. Plasma Sci.}
1387: \textbf{\bibinfo{volume}{20}}, \bibinfo{pages}{263} (\bibinfo{year}{1992}).
1388:
1389: \bibitem[{\citenamefont{{Nusinovich G.S., Vlasov A.N., Antonsen
1390: T.M.}}(2001)}]{Nusinovich:2001}
1391: \bibinfo{author}{\bibnamefont{{Nusinovich G.S., Vlasov A.N., Antonsen T.M.}}},
1392: \bibinfo{journal}{Phys.Rev.Lett.} \textbf{\bibinfo{volume}{87}},
1393: \bibinfo{pages}{218301} (\bibinfo{year}{2001}).
1394:
1395: \bibitem[{\citenamefont{{Dronov V., Hendrey M.R., Antonsen T.M., Ott
1396: E.}}(2004)}]{Dronov:2004_TWTChaos}
1397: \bibinfo{author}{\bibnamefont{{Dronov V., Hendrey M.R., Antonsen T.M., Ott
1398: E.}}}, \bibinfo{journal}{Chaos} \textbf{\bibinfo{volume}{14}},
1399: \bibinfo{pages}{30} (\bibinfo{year}{2004}).
1400:
1401: \bibitem[{\citenamefont{{Trubetskov D.I., Mchedlova E.S., Anfinogentov V.G.,
1402: Ponomorenko V.I., Ryskin N.M.}}(1)}]{Trubetskov:1996_CHAOS}
1403: \bibinfo{author}{\bibnamefont{{Trubetskov D.I., Mchedlova E.S., Anfinogentov
1404: V.G., Ponomorenko V.I., Ryskin N.M.}}}, \bibinfo{journal}{Chaos}
1405: \textbf{\bibinfo{volume}{6}}, \bibinfo{pages}{358} (\bibinfo{year}{1996}).
1406:
1407: \bibitem[{\citenamefont{{Trubetskov A.E., Hramov
1408: A.E.}}(2003)}]{TrueAeh:2003_microwave_electronics_1engl}
1409: \bibinfo{author}{\bibnamefont{{Trubetskov A.E., Hramov A.E.}}},
1410: \emph{\bibinfo{title}{Lectures on microwave electronics for physicists (In
1411: Russian)}} (\bibinfo{publisher}{Vol.~1. Fizmatlit, Moscow},
1412: \bibinfo{year}{2003}).
1413:
1414: \bibitem[{\citenamefont{{Packard~N.H., Crutchfield~J.P., Farmer~J.D.,
1415: Shaw~R.S.}}(1980)}]{Packard:1980_Delay}
1416: \bibinfo{author}{\bibnamefont{{Packard~N.H., Crutchfield~J.P., Farmer~J.D.,
1417: Shaw~R.S.}}}, \bibinfo{journal}{Phys. Rev. Lett.}
1418: \textbf{\bibinfo{volume}{45}}, \bibinfo{pages}{712} (\bibinfo{year}{1980}).
1419:
1420: \bibitem[{\citenamefont{{Takens F.}}(1981)}]{Takens:1981}
1421: \bibinfo{author}{\bibnamefont{{Takens F.}}}, in
1422: \emph{\bibinfo{booktitle}{Lectures Notes in Mathematics}}, edited by
1423: \bibinfo{editor}{\bibnamefont{{Rand D., Young L.--S}}}
1424: (\bibinfo{publisher}{N.Y.: Springler--Verlag}, \bibinfo{year}{1981}), p.
1425: \bibinfo{pages}{366}.
1426:
1427: \bibitem[{\citenamefont{{Schwarz U., Kurths J., Kliem B., Kr\"uger A., Urpo
1428: S.}}(1998)}]{Schwarz:1998_WaveletsAstro}
1429: \bibinfo{author}{\bibnamefont{{Schwarz U., Kurths J., Kliem B., Kr\"uger A.,
1430: Urpo S.}}}, \bibinfo{journal}{Astron. Astrophys., Suppl. Ser.}
1431: \textbf{\bibinfo{volume}{127}} (\bibinfo{year}{1998}).
1432:
1433: \bibitem[{Wav(2004)}]{WaveletsInPhysics:2004}
1434: \emph{\bibinfo{title}{Wavelets in Physics}}
1435: (\bibinfo{publisher}{Cambridge
1436: University Press}, \bibinfo{year}{2004}), \bibinfo{edition}{{J.C. Van den
1437: Berg}} ed.
1438:
1439: \end{thebibliography}
1440:
1441:
1442:
1443:
1444:
1445:
1446:
1447:
1448:
1449:
1450:
1451:
1452:
1453:
1454:
1455:
1456: \end{document}
1457: