Intermittent lag synchronization in a driven system of coupled oscillators

PRAMANA

— journal of

c Indian Academy of Sciences °

physics

Vol. 64, No. 4 April 2005 pp. 503–511

Intermittent lag synchronization in a driven system of coupled oscillators 2 ´ ALEXANDER N PISARCHIK1 and RIDER JAIMES-REATEGUI 1

Centro de Investigaciones en Optica, Loma del Bosque 115, Lomas del Campestre, Leon 37150, Guanajuato, Mexico 2 Universidad de Guadalajara, Campus Universitario Los Lagos, Enrique D´ıaz de Le´ on, Paseo de Las Monta˜ nas, Lagos del Moreno 47460, Jalisco, Mexico E-mail: [email protected] Abstract. We study intermittent lag synchronization in a system of two identical mutually coupled Duffing oscillators with parametric modulation in one of them. This phenomenon in a periodically forced system can be seen as intermittent jump from phase to lag synchronization, during which the chaotic trajectory visits a periodic orbit closely. We demonstrate different types of intermittent lag synchronizations, that occur in the vicinity of saddle-node bifurcations where the system changes its dynamical state, and characterize the simplest case of period-one intermittent lag synchronization. Keywords. Duffing oscillator; synchronization; chaos. PACS Nos 02.60.Cb; 05.45.Pq; 05.45.Xt

Synchronization of coupled oscillatory systems has attracted great interest in almost all areas of natural sciences, engineering and social life in the past few years because of its important practical applications which include communications, modelling brain and cardiac rhythm activity, earthquake dynamics, etc. (see, e.g., [1,2] and references therein). Different types of synchronizations such as complete synchronization [3], generalized synchronization [4], phase synchronization [5], lag synchronization [6], measure synchronization [7], almost synchronization [8], and anticipated synchronization [9] have been identified. Many of these theoretical findings have been experimentally verified in real systems, such as biological and medical [10] systems and chaotic lasers [11]. However, most of the theoretical and experimental works are devoted to investigation of synchronization effects in self-oscillatory autonomous systems. Significantly less attention has been given to a study of synchronization of driven coupled oscillators, in which either a system parameter or a state variable is periodically modulated. One of the most extensively investigated models of nonlinear systems with external forcing is the Duffing oscillator. The Duffing oscillator was successfully explored to model a variety of physical processes such as stiffening strings, beam buckling, nonlinear electronic circuits, superconducting Josephson parametric

503

Alexander N Pisarchik and Rider Jaimes-Re´ ategui amplifiers, and ionization waves in plasmas, as well as biological and medical processes. For example, the transition to hyperchaos and synchronization phenomena in a system of coupled Duffing oscillators were investigated respectively by Kapitaniak [12] and Landa and Rosenblum [13]. The bifurcation structure of two coupled periodically driven Duffing oscillators in space of modulation parameters was studied by KozÃlowski et al [14] for the case of single-well potentials and by Kenfack [15] for the case of double-well potentials. The effect of phase difference in mutually coupled chaotic oscillators was considered by Yin et al [16]. Recently, Raj et al [17] investigated coexisting attractors and synchronization of chaos in two coupled Duffing oscillators with two driving forces. Usually, an external driving force is applied to a state variable in one of the oscillators or to variables of both oscillators. However, in a real experimental practice it is more convenient to modulate a system parameter rather than a state variable. The parametric modulation is commonly used in electromechanical and electronic systems, in particular, for communication purpose. Nevertheless, only few works were devoted to a study of synchronization of parametrically modulated systems [18,19]. Recently, phase synchronization of chaotic oscillations has been found by Rosenblum et al [5] in autonomous non-identical oscillators with symmetric coupling. This regime is characterized by a perfect locking of the phases of the two signals, while the two chaotic amplitudes remain uncorrelated. Later, the same authors observed lag synchronization, that consists of hooking one system to the output of the other shifted in time of a lag time τlag [s1 (t) = s2 (t − τlag )] [6]. The latter phenomenon has been observed experimentally in two unidirectionally coupled Chua’s circuits [20]. The present research has been stimulated by the work of Boccaletti and Valladares [21] who characterized intermittent lag synchronization (ILS) of two nonidentical symmetrically coupled Rössler systems. They observed intermittent bursts away from the lag synchronization to asynchronous regime and described this phen (n = 1, 2, ...), such nomenon in terms of the existence of a set of lag times τlag n that the system always obeys s1 (t) ' s2 (t − τlag ) for a given n. In this work we demonstrate a similar behavior in a driven system of coupled oscillators. Unlike a self-oscillatory system, all oscillations in our system are the forced oscillations induced and driven by the external periodic forcing and hence they are always phase-locked with the forcing, and so, in a driven system ILS manifests itself as an intermittent behavior between phase and lag synchronization. Generally, we consider a system of two coupled subsystems: x˙ = g(x, y; A) and y˙ = h(x, y; B), where x and y are phase-space variables, A and B are sets of parameters, and g and h are the corresponding nonlinear velocity fields. If one of the parameters, say a (a ∈ A), is a function of time, i.e., a = ϕ(t) while the other parameters are constants, the two subsystems are not completely identical, i.e. g 6= h. Nevertheless, we may consider the subsystems to be almost identical when averaged in time parameters are the same, i.e., hAi = B(h· · ·i denotes temporal average). Dynamics of two identical nonlinear oscillators can be governed by the equation dV (x) , (1) dx where x ≡ (x, y), γ is a damping factor, and V (x) is a two-dimensional anharmonic ¨ + γ x˙ = − x

504

Pramana – J. Phys., Vol. 64, No. 4, April 2005

Intermittent lag synchronization potential function of the coupled oscillators. The potential functions for symmetric Duffing oscillators can be expressed as follows: a 2 b 4 δ 2 2 x + x + x y , 2 4 2 a0 2 b 4 δ 2 2 V (y, x) = y + y + x y , 2 4 2

V (x, y) =

(2) (3)

where a, a0 and b are parameters and δ is a coupling coefficient. Without external modulation, eqs (1)–(3) do not have periodic solutions. The external modulation is added to one of the oscillators in the form of parametric modulation a = a0 [1 − m sin(2πf t)],

(4)

where m and f are the modulation depth and frequency. Here we consider only the case of a double-well potential, i.e., when a0 < 0, with positive b (b > 0). This case is more useful for modelling a real experiment for signal transmission, because the oscillators have non-zero stable equilibrium points unlike a single-well case. Equations (1)–(4) can be written as x˙ 1 x˙ 2 x˙ 3 x˙ 4

= x2 , = −γx2 − a0 [1 − m sin(2πf t)]x1 − bx31 − δx1 x23 , = x4 , = −γx4 − a0 x3 − bx33 − δx21 x3 .

(5) (6) (7) (8)

The general information analysis of dynamical regimes, which can be expected from eqs (5)–(8) with parameters γ = 0.4, a0 = −0.25, and b = 0.5, reveals the following possible situations: (i) When both coupling strength and modulation amplitude are sufficiently small (δ . 0.1 and m . 0.1), the mismatch of natural frequencies of the two oscillators is also small and the bifurcation diagrams of the variables of each subsystem (x1 and x3 ) have the standard shape of a linear resonance response (figure 1a), i.e., the two oscillators are completely synchronized. (ii) When δ increases, the response becomes nonlinear and both resonance frequencies move to a higher frequency region and the mismatch also increases (figure 1b). Thus, the oscillations occur to be shifted in time, i.e., lag synchronization takes place. For δ < 0.5 and m < 0.5, both subsystems oscillate in a periodic regime with f (period 1) over all frequency range. Finally, when δ → 0.5 the nonlinear resonances disappear in the system response (figure 1c). (iii) A further increase in δ leads to the appearance of coexisting multiple periodic attractors and steady-state solutions (figure 1d) [19]. Although the two oscillators are almost identical, the origin of the lag in their oscillations is the same as in the case of non-identical autonomous oscillators [6,21], namely, a mismatch of their nonlinear resonance frequencies, that appears due to the nonlinear coupling and because the modulation is applied only to one of the oscillators. In this paper we are interested mainly in a chaotic region. Chaotic oscillations in the system (eqs (5)–(8)) are observed at relatively high modulation amplitudes (m > 0.75) and low couplings (δ < 0.25). In parametrically modulated coupled systems, three parameters (δ, m, and f ) can be used as control parameters, and hence Pramana – J. Phys., Vol. 64, No. 4, April 2005

505

Alexander N Pisarchik and Rider Jaimes-Re´ ategui

Figure 1. Bifurcation diagrams for driven (x1 ) and passive (x3 ) oscillators with respect to modulation frequency for m = 0.1 for different coupling strengths (a) δ = 0.1, (b) 0.4, (c) 0.5, and (d) 0.51.

the dynamics may be analysed in the space of these three parameters. In figure 2 we present the co-dimensional-two bifurcation diagrams in the (δ, m) (figure 2a) and (m, f ) (figure 2b) parameter spaces. The saddle-node bifurcation (SNB) lines bound different dynamical regimes: periodic orbits (PO), one-well chaos (OWC), cross-well chaos (CWC), and hopping oscillations (HO) (periodic windows). As mentioned above, all oscillations in our system are excited by the external periodic modulation and therefore they are always phase-locked with the forcing. In a periodic regime, the state variables of the two subsystems are shifted in time, i.e., they are lag synchronized. Within very narrow parameter range, close to SNBs, short windows of periodicity appear intermittently in chaotic time series. In such an intermittent regime, the system occasionally jumps from chaos to local periodicity (figures 3 and 4). During these jumps, the chaotic trajectory visits a periodic orbit closely, i.e., we have a sort of intermittent lag synchronization observed in the driven system. The regime of ILS appears in regions of the parameter space in the vicinity of SNBs and is associated with on–off intermittency [22]. This type of intermittency (in other words, modulational intermittency) appears only in a driven system in which the external forcing is applied in the form of either noise, or chaos, or periodic modulation. On–off intermittency has been also detected experimentally in different dynamical systems [23–27].

506


Intermittent lag synchronization

Figure 2. Co-dimensional-two bifurcation diagrams in parameter spaces of (a) coupling strength δ and modulation depth m for f = 0.1 and (b) modulation frequency f and depth m for δ = 0.1. Intermittent lag synchronization occurs in the vicinity of the saddle-node bifurcation lines which bound different dynamical regimes: one-well chaos (OWC), cross-well chaos (CWC), hopping oscillations (HO), and periodic orbits (PO). The dots indicate the parameters for which the regimes of period-1 ILS (P1) and period-2 ILS (P2) shown in figures 3 and 4 are observed.

In figures 3 and 4 we demonstrate two kinds of ILS in the driven system: onestate period-1 (P1) ILS (figure 3) and cross-state period-2 (P2) ILS (figure 4). In the former case, the x1 -trajectory jumps intermittently from cross-well chaos to the small P1 orbit around each of the potential well and back. In the latter case, the trajectory jumps from cross-well chaos to the large P2 orbit oscillating between the two wells. Figures 3b and 4b display the enlarged parts of the time series where lag synchronization is observed. The regimes shown in figures 3 and 4 are observed for the parameters marked in figure 2 by the dots. These dots lie on the SNB lines which bound respectively the one-well and cross-well chaotic regions and the regions of hopping oscillations and cross-well chaos. Similar regimes of ILS are found near the other SNB lines. Recently, Rosenblum et al [6] proposed to describe the occurrence of ILS as a situation where during some periods of time the system verifies ∆ ≡ |x3 (t) − x1 (t − τ )| ¿ 1 (τ being a lag time), but where bursts of a local non-synchronous behavior


507

Alexander N Pisarchik and Rider Jaimes-Re´ ategui

Figure 3. One-state period-1 intermittent lag synchronization of one-well chaos (small-orbit synchronization). (a) Time series of driven (x1 ) and passive (x3 ) oscillators, (b) enlarged part of (a) demonstrating synchronous unstable period-1 orbits around the potential wells. δ = 0.1, f = 0.107, m = 0.8.

Figure 4. Two-state period-2 intermittent lag synchronization of cross-well chaos (large-orbit synchronization) in x direction. (a) Time series of driven (x1 ) and passive (x3 ) oscillators, (b) enlarged part of (a) demonstrating synchronous unstable period-2 orbits. δ = 0.1, f = 0.087, m = 0.8.

may occur. They identified this phenomenon with on–off intermittency and the bursts from lag synchronization were found to result from a small, but negative value of the second global Lyapunov exponent of the system, so that the trajectory visits attractor regions where the local Lyapunov exponent is still positive. In periodically driven systems, this condition should be modified to satisfy ∆ ≡ |x3(t) − hx3(t)i − η [x1 (t − τ ) − hx1 (t)i]| ¿ 1, (9) ¢ ¢ ¡ max ¡ max min min is a proportionality coefficient between the where η = x3 − x3 / x1 − x1 alternative amplitudes of the variables in the synchronous regime. The coefficient 508


Intermittent lag synchronization

Figure 5. Time series of ∆0 (τ0 = 116) in period-1 intermittent lag synchronization regime. The windows with ∆0 ≈ 0 can be viewed as the low-dimensional ‘lag synchronous’ attractor. (b) Similarity function S 2 (τ ) vs. lag time τ . There exists a global minimum at τ0 = 116 and local minima for smaller and larger times τn (n = 1, 2, 3, ...).

η is introduced because the modulation is applied only to one of the oscillators (x 1 ) and hence x3 < x1 . Therefore, the averages of the two variables are also different and hence they should be normalized. Of course, the criterion (eq. (9)) can be used only for characterization of the simplest case of P1 ILS shown in figure 3. For higher periods (P2, P3,...) of ILS, the shapes of the oscillations in the periodic windows are different for two oscillators, and hence more complex relationship is required. The temporal behavior of ∆0 (τ0 = 116) for the case of P1 ILS is shown in figure 5a. If the criterion eq. (9) is correct, the function ∆0 should be approximately equal to zero in the windows of the lag synchronous regime, as seen in figure 5a. Similarly to Rosenblum et al. [6], we may characterize lag synchronization by similarity function S(τ ), defined as time averaged difference ∆, conveniently normalized to the geometrical average of the two mean signals 2® ∆ 2 S (τ ) = , (10) 2 [hx1 (t)ihx23 (t)i]1/2 and search for its global minimum σ = minτ S(τ ), for τ0 6= 0. The similarity function vs. the lag time shown in figure 5b resembles the dependence similar to Pramana – J. Phys., Vol. 64, No. 4, April 2005

509

Alexander N Pisarchik and Rider Jaimes-Re´ ategui those reported previously by Boccaletti and Valladares [21] for two coupled Rössler systems. Looking at figure 5b, one can see that, besides a global minimum at τ0 = 116, S (τ ) displays many other local minima at smaller and larger lag times τn (n = 1, 2, 3, ...). The depth of the nth local minimum is closely related to the fraction of time when the corresponding lag configuration is closely visited by the system. The different lag times τn can be expressed by the relation τn ≈ τ0 + nT , where T is the period of external modulation or the return time of the limit cycle onto the Poincaré section. The anharmonicity in function S 2 (τ ) results from the anharmonicity of the periodic oscillations due to high nonlinearity of the system. In summary, we have studied synchronization properties of two mutually coupled Duffing oscillators with parametric modulation in one of them and have found synchronous states, which we identify with intermittent lag synchronization. In the intermittent states, the system during its temporal evolution occasionally changes its behavior from phase synchronization to lag synchronization. The regime of intermittent lag synchronization appears in regions of the parameter space close to saddle-node bifurcations and is associated with on–off intermittency. Recently, methods of closed-loop control [28] and open-loop control [29] of this phenomenon have been suggested. We believe that the main features of the synchronization phenomena obtained in the coupled Duffing oscillators are common for a wide class of parametrically driven systems and may be observed in experiments. Acknowledgements ANP thanks S Boccaletti for valuable discussions. This work has been supported through a grant from the Institute Mexico-USA of the University of California (UC-MEXUS) and Consejo Nacional de Ciencia y Tecnologia (CONACyT).

References [1] A Pikovsky, M Rosenblum and J Kurths, Synchronization: A universal concept in nonlinear science (Cambridge University Press, New York, 2001) [2] S Boccaletti, J Kurths, G Osipov, D L Valladares and C Zhou, Phys. Rep. 366, 1 (2002) [3] H Fujisaka and T Yamada, Prog. Theor. Phys. 69, 32 (1983) [4] N F Rulkov, M M Sushchik, L S Tsimring and H D I Abarbanel, Phys. Rev. E51, 980 (1995) [5] M G Rosenblum, A S Pikovsky and J Kurths, Phys. Rev. Lett. 76, 1804 (1996) [6] M G Rosenblum, A S Pikovsky and J Kurths, Phys. Rev. Lett. 78, 4193 (1997) [7] A Hampton and H D Zanette, Phys. Rev. Lett. 83, 2129 (1999) [8] R Femat and G Solis-Perales, Phys. Lett. A262, 50 (1999) [9] H U Voss, Phys. Rev. E61, 5115 (2000) [10] C Schafer, M G Rosenblum, J Kurths and H H Abel, Nature (London) 392, 239 (1998) P Tass, M G Rosenblum, J Weule, J Kurths, A Pikovsky, J Volkmann, A Schnitzler and H J Freund, Phys. Rev. Lett. 81, 3291 (1998) A Neiman, X Pei, D Russell, W Wojtenek, L Wilkens, F Moss, H A Braun, M T Huber and K Voigt, Phys. Rev. Lett. 82, 660 (1999)

510


Intermittent lag synchronization [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23]

[24] [25] [26] [27] [28] [29]

G D Van Wiggeren and R Roy, Science 279, 1198 (1998) T Kapitaniak, Phys. Rev. E47, R2975 (1993) P S Landa and M G Rosenblum, Appl. Mech. Rev. 46, 141 (1993) J KozÃlowski, U Parlitz and W Lauterborn, Phys. Rev. E51, 1861 (1995) A Kenfack, arXiv:nlin.CD/0304022, v.1, 14 Apr. 2003 H-W Yin, J-H Dai and H-J Zhang, Phys. Rev. E58, 5683 (1998) S Paul Raj, S Rajasekar and K Murali, Phys. Lett. A264, 283 (1999) B F Kuntsevich and A N Pisarchik, Phys. Rev. E64, 046221 (2001) A N Pisarchik, Phys. Lett. A318, 65 (2003) L Zhu and Y-C Lai, Phys. Rev. E64, 045205(R) (2001) S Boccaletti and D L Valladares, Phys. Rev. E62, 7497 (2000) E A Spiedel, Ann. NY Acad. Sci. 617, 305 (1981) T Yamada, K Fukushima and T Yazaki, Prog. Theor. Phys. Suppl. 99, 120 (1989) P W Hammer, N Platt, S M Hammel, J F Heagy and B D Lee, Phys. Rev. Lett. 73, 1095 (1994) D L Feng, C X Yu, J L Xie and W X Ding, Phys. Rev. E58, 3678 (1998) F R¨ odelsperger, A Cenys and H Benner, Phys. Rev. Lett. 75, 2594 (1995) T John, R Stannarius and U Behn, Phys. Rev. Lett. 83, 749 (1999) A N Pisarchik and V J Pinto-Robledo, Phys. Rev. E66, 027203 (2002) Y Nagai, X-D Hua and Y-C Lai, Phys. Rev. E54, 1190 (1996) R J Reategui and A N Pisarchik, Phys. Rev. E69, 067203 (2004)


511