Medium access control (MAC) design for ad hoc networks is studied in this paper based on game theory. We propose the noncooperative random access game, where the link's utility function captures its demand for successful transmission and the pricing of persistence probability is used to improve efficiency for medium access. First, we consider a pricing function that is a linear function of the persistence probability. This pricing function leads to a simple iterative scheme by which each link can arrive at the Nash equilibrium (NE) using only its local measurements. The existence, uniqueness and stability of the NE for the game are proved. Since the persistence probability at equilibrium point, the system throughput and the packet collision probability all depend on the fixed price coefficient, we propose the adaptive price setting scheme based on the measurement for network conditions. The adaptive pricing scheme admits a Pareto dominant NE and enables higher link rates with lower persistence probabilities. Simulation results illustrate that the fast convergence, fair and efficient resource allocation of these schemes can be achieved with less collision compared with the standard binary exponential backoff protocols.
展开▼
