Network Cost Minimization Leads to r/K-Selection in the Computer Networks

Today´s Internet hosts a wide variety of network objects (namely, netOs) like contents, resources and functionalities and so on, which are often decomposed and distributed over the Internet to provide and support network-based services and applications conveniently. However, it is a big challenge to efficient management due to the huge number of existing netOs. Based on our investigations, most existing works that organizing those netOs in the computer network efficiently formulates network utility maximization problems (or network cost minimization). Although some works are effective in the corresponding aspects, the Internet still suffers from inefficiency and ossification. Instead of solving a concrete problem, we make the first attempt to model the numerical evolution of netOs (namely, EvoNetO) to uncover the hidden order of netOs. We prove that the problem of minimizing network cost is equivalent to fastest decay rate or slowest growth rate in the population dynamics. And then we find that the netOs distributed over the network will go to two extreme, i.e., r-strategists and K-strategists under the goal of minimizing network cost for netO placements. This phenomenon exists widely in the computer network.