Synchronization and Complex Dynamics of Oscillators with Delayed Pulse Coupling
Synchronization and Complex Dynamics of Oscillators with Delayed Pulse Coupling
复制标题
DOI:
10.1002/anie.201205214
复制
发表时间:
2012-01-01
影响因子:
16.6
通讯作者:
Torcini, Alessandro
中科院分区:
文献类型:
--
作者:
Baer, Markus;Schoell, Eckehard;Torcini, Alessandro
Networks of coupled nonlinear oscillators often exhibit synchronization and complex dynamic patterns that range from oscillator clusters and partially synchronized states to wave patterns and spatiotemporal chaos. The emerging behavior depends crucially on the type and strength of the coupling, the network topology, as well as on the frequency distributions of the oscillators in the network. The relevance of synchronization phenomena has been recognized for many important biological functions, and the pioneering studies of Winfree and Kuramoto has opened up a challenging field of mathematical and theoretical study.[1] Chemical systems such as the Belousov–Zhabotinsky (BZ) reaction or coupled electrochemical oscillators have previously been used to study synchronization and related phenomena, such as the formation of oscillatory clusters, experimentally. In the BZ system, the oscillators in the form of catalyst particles in a catalyst-free background [2a] or microfluidic water droplets embedded in a nonreactive oil phase [2b] are coupled diffusively, whereas electrochemical systems [2c–e] constitute a realization of global all-to-all coupling of the involved oscillators. These systems proved to be flexible test beds for synchronization studies, because the number of oscillators (10–100000) as well as the geometrical arrangement of the oscillators can be easily varied. A further step in this success story was recently provided by Horvath et al.,[3] who investigated a pair of pulse-coupled oscillators, with delay realized by continuously fed stirred tank reactors (CSTRs) filled with an oscillatory BZ mixture and coupled to a controlled sudden release of a chemical activator (AgNO3) or inhibitor (BrÀ) of the BZ reaction. The reactors are operated in a regime of spiking oscillations, and coupling occurs through the release of a chemical agent into one of the CSTRs only if a spike is recorded in the other CSTR. By changing the amount of the released substance Horvath etal. could vary the coupling strength, and by controlling the time of release (= time of the pulse coupling) they were able to introduce an arbitrary time delay. A detailed chemical kinetics model provides, in addition to the experiments, a convincing mechanistic explanation of the observed behavior.[3]Their study is highly relevant not only for nonlinear chemical dynamics but also for important fields of biological research, since most biological oscillators—from cardiac pacemakers to flashing fireflies—are, in fact, pulse coupled. This fact motivated a large number of mathematical studies and predictions of their behavior, such as the proof by Mirollo and Strogatz that an arbitrary number of oscillators with excitatory all-to-all pulse coupling synchronize.[4] The importance of a time delay in the coupling has been recognized recently in a number of high-profile experimental studies in biology ranging from synthetic genetic oscillators [5a] to the segmentation clock in the embryonic development of vertebrae.[5] An important field where pulse-coupled oscillators with delay are widely discussed is neuroscience. In this context, experiments have been devoted to the investigation of two coupled neurons,[6a] of the synchronous activity of different regions of the brain,[6b] as well as of the emergence of b or g oscillations in large neuronal populations, which are often detected by electroencephalography.[6c] The importance of such experiments has resulted in numerous computational studies on delayed pulse-coupled oscillators in neuroscience. For pairs of oscillators, it was discovered that the presence of delayed inhibitory coupling leads to synchronization more reliably than does excitatory coupling.[7a] Other …