Incentive-oriented downlink scheduling for wireless networks with real-time and non-real-time flows

We study the problem of designing downlink scheduling policy for wireless networks with both real-time and non-real-time flows. Most existing scheduling policies for these networks unfairly favor one type of flows. As a result, strategic clients may lie about their types of traffic in order to get more service from the base station. To counter these strategic behaviors, a scheduling policy should be incentive compatible in the sense that clients cannot improve their performance by lying. We propose a novel scheduling policy that is incentive compatible. The policy can be easily implemented as a thin layer between the Network layer and the MAC layer. It can also work with a wide range of upper-layer and lower-layer protocols while providing satisfactory performance to both real-time and non-real-time flows. Simulation results show that this policy outperforms other widely used policies by large margins under a variety of scenarios.