In many animals and insects, hearing is very acute to the faintest of sounds; the underlying mechanism can be explained by self-tuning. Recently, signal response amplification has been shown to be implemented through networks exhibiting scale-free topology, which has potential applications in artificial information processing systems and devices. We review in this paper the main results obtained in networked double-well oscillators and briefly discuss future research directions.signal response, scale-free network, self-tuning, double-well oscillator
Citation:Liu Z H. Signal response amplification of scale-free networks. Chinese Sci Bull, 2011Bull, , 56: 3623-3629, doi: 10.1007 Systems that can detect and amplify signals at specific frequencies are commonplace in the natural world and most notably in the visual and auditory systems of animals [1]. Signal detection in animals is through light-and auralsensitive organs, constituted by a large number of networked units. For example, cells in living organisms respond to their environment by an interconnected network of receptors, messengers, protein kinases and other signaling molecules [2][3][4]. One of the more prominent features of our hearing system is the ability to perceive sound stimuli that range over six orders of magnitude in sound pressure [5]. Hair cells within the cochlear are stimulated by sound waves, the induced motion being amplified at characteristic locations that depends on the frequencies of sound. These cells transmit signals to the auditory nerve [6]. It is well known that many animals and insects have the ability to detect faint sounds from their environment. Physiological evidence exists for a range of animals and insect auditory systems that this active audition is due to Hopf bifurcations [7][8][9]. Models have also been proposed to develop the underlying mechanism behind enhanced amplification in hearing systems, i.e. self-tuned critical oscillations of hair cells nearby the Hopf bifurcation. 2 , where a controls the barrier height of the potential; x=±1 are the locations of the two minima. Suppose an oscillator has probability w + (w ) of jumping from the right (left) well to the left (right) well; then self-tuning means that the parameter a can be self-adjusted by the following equation:
a t a t w t p t w t p t twhere p + (p ) denotes the probability that the oscillator stays at x=±1. If w + and w are small, the first term in eq.(1) will be larger than the second term and thus result in a decrease in a. For larger w + and w , the first term in eq.(1) will be smaller than the second term and thus result in an increase in a. Therefore, the parameter a is self-tuned to an optimal value by switching probabilities w + and w . Scientists and engineers frequently take inspiration from