SGER: Connection Admission Control and Network Pricing in Public Safety Spectrum Sharing

该项目正在解决以下问题：(1)什么样的动态频谱共享机制能为两种类型的用户提供最大的利益？公共安全用户抢先使用频谱可能会给商业用户带来不必要的成本，而商业用户的连接会被切断。在另一种极端情况下，无论调用方类型如何，只要可用就分配通道的简单共享可能会对公共安全造成过高的阻塞。许多中间机制可以公式化。这些机制都可以与目前将总带宽严格划分为公共安全和商业子频段的机制进行比较。(2)从公共安全用户的封锁成本和商业用户的经济成本来看，选择这些不同机制的成本是多少？该项目正在开发能够更精确地描述共享机制的好处的方法和更优化共享机制的技术。利用强化学习和网络定价的方法被应用于开发最优共享机制的问题。这项工作将有助于量化动态频谱共享的好处，并向公共安全和商业用户保证共享概念将满足他们各自的需求。

This project is addressing the following questions:(1) What dynamic spectrum sharing mechanisms provide the most benefit to both types of users? Preemptive spectrum use by public safety users may incur undue cost to commercial users which have their connections cut. At the other extreme, simple sharing that allocates channels whenever available regardless of caller type may incur too high a blocking to public safety. Many intermediate mechanisms can be formulated. These mechanisms can all be compared with the current mechanism that divides the total bandwidth into rigidly segregated public safety and commercial sub-bands. (2) What is the cost of choosing these different mechanisms in terms of blocking for public safety users and economic cost to commercial users?This project is developing methods that will allow a more precise characterization of the benefit of sharing mechanisms and techniques for more optimal sharing mechanisms. Methods that draw on reinforcement learning and network pricing are being applied to the problem of developing optimal sharing mechanisms. This effort will help quantify the benefit of dynamic spectrum sharing and provide assurances to public safety and commercial users that the sharing concept will meet their separate needs.