• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>IEEE Conference on Local Computer Networks >Q-Learning Based Co-Operative Spectrum Mobility in Cognitive Radio Networks
【24h】

Q-Learning Based Co-Operative Spectrum Mobility in Cognitive Radio Networks

机译：认知无线电网络中基于Q学习的协作频谱移动性

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

In cognitive radio systems, fast and efficient spectrum selection is a vital task to minimize the overhead of spectrum scanning, and hence to improve the response time of the system. So, the choice of channel sensing sequence plays an important role for better performance of the system. This paper proposes a co-operative Q-learning based spectrum sensing technique for the secondary users of an ad hoc network to access the primary channels. By the proposed technique, every secondary user (SU) maintains a dynamic priority list of channels based on Q-learning from its own action-observation history, as well as from spatial channel information exchange among its local neighbors. Whenever there is a demand an SU scans the spectrum according to the order in the priority list until there is a success. Simulation studies show that with significantly less computing and scanning overhead, our proposed Q-learning based approach improves the response time and call block/drop rate to offer better performance compared to other contemporary reinforcement learning based approaches.
机译：在认知无线电系统中，快速有效的频谱选择是至关重要的任务，以最大程度地减少频谱扫描的开销，从而改善系统的响应时间。因此，信道感测序列的选择对于系统的更好性能起着重要的作用。本文提出了一种基于协作Q学习的频谱感知技术，供自组织网络的次要用户访问主信道。通过提出的技术，每个辅助用户（SU）都根据从其自己的动作观察历史以及其本地邻居之间的空间信道信息交换中的Q学习来维护信道的动态优先级列表。只要有需求，SU就会根据优先级列表中的顺序扫描频谱，直到成功为止。仿真研究表明，与其他当代的基于增强学习的方法相比，我们提出的基于Q学习的方法显着减少了计算和扫描开销，从而改善了响应时间和呼叫阻塞/掉线率，从而提供了更好的性能。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号