Research On Complex Resource Allocation System With Network Structure

In both natural and social systems, there exist various kinds of competition for limited resources. However, for the purposes of theoretical analysis, past studies usually focus on the competition in which the interactions between the individuals are global. In real world, the individuals only compete for the resources with some of the other individuals i.e., compete locally. Moreover, the network structure between the individuals is worth considering. In this paper, we study the system in which the individuals can only compete locally through human experiment and computer simulation.Firstly, we use random network to model the competition relationship between the individuals in a resource allocation system and study how the different sizes of the clusters would impact the properties of the whole system. We find that even if the individuals could only compete locally, the resource allocation may also be efficient. Moreover, a certain size of the cluster may also increase the system stability. We also find if only local competition was allowed, there would exist some steady arbitrage opportunities. This phenomenon inspires us to study the evolution of the macro states of the system reference the principle of increasing entropy in the statistical physics. We find that there exists non-equilibrium steady states in the isolated human system. The increasing adaptability of the human subjects is exactly the reason that the entropy of the system would increase. And the local competition limits the increasing of the adaptability so that the non-equilibrium steady states could exist.At last, we reports the efficiency of the resource allocation system under other network structures.