CAJ | 학술논문

下一代无线网络是多种无线接入技术并存的异构网络，要充分利用各种无线技术网络的资源，为用户提供更好的服务，需要实现异构网络的融合。本文引入认知无线网络概念以解决未来网络的复杂性、异构性问题，从而实现异构网络的融合。认知无线网络能够使网络从静态工作模式发展到动态自适应工作模式，从单一封闭式网络发展到异构融合网络，从而能够面向用户需求提供更加可靠和高速率的用户服务。无线接入网络的选择问题是未来无线异构网络的关键技术之一，也是多网融合的基础。本文面向用户终端提供一种基于 Q-学习的认知无线网络接入方案，该方法充分利用了认知无线电技术以及认知无线网络相关技术，使得移动终端能够根据自身的偏好或者QoS需求，选择最合适的接入网络，同时实现系统整体能耗开销最优化。特别地，本文以家庭基站和宏基站共存复杂架构下的网络接入问题进行探讨并提供解决办法。仿真结果表明，本文提供的网络选择算法能够显著提升系统的阻塞情况并降低系统的能耗开销。
하일대무선망락시다충무선접입기술병존적이구망락，요충분이용각충무선기술망락적자원，위용호제공경호적복무，수요실현이구망락적융합。본문인입인지무선망락개념이해결미래망락적복잡성、이구성문제，종이실현이구망락적융합。인지무선망락능구사망락종정태공작모식발전도동태자괄응공작모식，종단일봉폐식망락발전도이구융합망락，종이능구면향용호수구제공경가가고화고속솔적용호복무。무선접입망락적선택문제시미래무선이구망락적관건기술지일，야시다망융합적기출。본문면향용호종단제공일충기우 Q-학습적인지무선망락접입방안，해방법충분이용료인지무선전기술이급인지무선망락상관기술，사득이동종단능구근거자신적편호혹자QoS수구，선택최합괄적접입망락，동시실현계통정체능모개소최우화。특별지，본문이가정기참화굉기참공존복잡가구하적망락접입문제진행탐토병제공해결판법。방진결과표명，본문제공적망락선택산법능구현저제승계통적조새정황병강저계통적능모개소。
The future network is a network coexisted of multiple heterogeneous wireless access technologies. It is required that various wireless technologies be integrated to make full use of network resources and to provide users with better services. Thus there is urgent need to achieve the integration of heterogeneous networks. In the paper, the concept of cognitive wireless networks is therefore introduced to solve the complexity and heterogeniety of future networks. It transfers the heterogeneousthe network from conventionall static mode to dynamic and adaptive mode, from multiple single closed-end networks to a converged heterogeneous network, which provides more reliable and high-speed services. In the converged heterogeneous wireless network, access network selection is one of the key technologies of the next generation wireless networks and is aslo the basis of multi-network integration. This article is intended to provide a user terminal Q-learning based on cognitive wireless network access solutions. The method makes full use of cognitive radio and cognitive wireless network-related technology, making the mobile terminal select the most appropriate network to access according to their own preferences or QoS requirements, while leading to overall power consumption optimization. In particular, access network selection issue under the complex architecture of macro/femtocell coexistence scenario is discussed and resolved. Simulation results show that the provided network selection algorithm offers significant blockage enhancements to the system and reduces the overall power consumption.