Proposals and simulations from brain science for regulating large-scale and wide-spread information systems

点击翻译按钮获取中文摘要

This study explores the way of coexistence of human being and information system by investigating biological information systems such as brain and biosystems.Simulations were performed using interconnected neural network models consisting of spiking neurons, where neuronal spikes could be regarded as an information carrier in the brain (packets in a computer network). Resulting cluster patterns of neuronal activities are similar to those of packet flow in the computer network. In addition, enhancing mutual connections between neurons augments the clustering. This result indicates that enhancing inter-node connections could result in localizing highly activated nodes in the computer network.Possible control strategy employed in cardiovascular system is explored through modeling, which underlies 1/f fluctuations in heart rate and blood pressure. The results clarified that the control strategy is not a rigid one prohibiting deviations of control variables from an operational point, but that it is a loose one allowing the variations to some extent. Simultaneously, the hierarchical control strategy is shown to be essential that switches global activation level of the control variable.Resulting dynamics of hybrid system coupling an actual hippocampal neural network and nonlinear oscillator is investigated for varied coupling mode. This study suggests that a system of essentially high degree of freedom happens to exhibit behavior of lower degree of freedom dependent on control strategy. Brain is considered to realize such a diverse dynamics in the neural network by varying controls.Thus clarified control strategies employed in the brain and biosystem could provide a novel framework for network control and information access control in the internet.