Attractor selection and perturbation for robust networks in fluctuating environments

The heterogeneity of connected devices, technologies, and protocols anticipated in future-generation information and communication networks also requires the development of new approaches for robust and self-adaptive systems. Recently, methods observed from biological phenomena have gained much attention as viable alternatives or inspiration for the solution of networking problems. The main advantages of such dynamic mechanisms inspired by biology lies in their self-organizing properties, scalability to numbers of connected devices, simplicity in terms of control rules, as well as adaptability and robustness to changing and fluctuating environments. Dealing with fluctuations plays a key role in maintaining the stability of the system. In this article we propose a new framework for selecting among different networks and services based on the robustness of each network¿s performance metrics. The selection is only based on observations of the system's responsiveness to inherent fluctuations. This method is derived from biological experiments where the speed of fluorescence evolution of proteins in bacteria is observed to have a positive correlation with the phenotypic fluctuation of fluorescence over clone bacteria. Due to the explicit utilization of the inherent fluctuations in the system, the proposed selection scheme can operate smoothly to select the most suitable and robust network.