基于运动预测的ABC支持型资源预约机制

Motion Prediction Based Resource Reservation Mechanism with ABC Support

面向下一代互联网NGI(Next Generation Internet)需支持服务质量QoS(Quality of Service)和总最佳链接ABC(Always Best Connected)的实际需求,设计了一种基于运动预测的ABC支持型资源预约机制。引入效用作为"最佳"的度量,综合描述了QoS满意度、网络提供商偏好、网络编码类型偏好、网络覆盖范围偏好、网络耗电程度等因素。利用用户历史运动信息和当前运动信息进行预测,在满足条件的接入网络中预约资源,同时引入资源借用策略来提高资源预约成功率。仿真结果表明,该机制是有效的,性能较好。

In order to support ABC （Always Best Connected） and provide QoS （Quality of Service） in NGI （Next Generation Interne0, a mechanism named motion prediction based resource reservation mechanism with ABC support was proposed. Utility was used as the measure to ＇best＇, which comprehensively reflected QoS satisfaction degrees, network provider preferences, network coding type preferences, network coverage preferences, power consumption levels, etc. Both historical and instant movement information was applied to motion prediction, and then resource reservation was done in the access network under suitable conditions. At the same time, the resource borrowing strategy was adopted to improve the resource reservation success rate further. Simulation results have shown that this proposed mechanism is both feasible and effective.