| Prefetching has been shown to be an effective technique for reducing resource cost and delay in heterogeneous wireless networks. However, in modern wireless local area networks, there is little centralized management, with no control of upper-level functions such as prefetching, and so users are free to behave selfishly. This work focuses on how pricing can be used to control the suboptimality that results from prefetching and selfish users in heterogeneous wireless networks, and how the perceived cost for the user can be optimized. We derive an analytic model to characterize the optimal network and Nash Equilibrium prefetching strategies. We present a pricing scheme that optimizes the best achievable perceived cost when the network is in a Nash Equilibrium. Finally, we study the effects of network load, the value of time, and user mobility on the optimal pricing scheme. |
