CAJ | 학술논문

为了避免多媒体多播组播单频网(MBSFN)区域内部和区域之间的干扰，进一步提高频谱效率，该文提出一种改进的基于噪声调节时滞噪声混沌神经网络(NHNCNN)的动态信道分配方法。首先，根据MBSFN区域的特殊拓扑结构，重新定义了4种电磁兼容限制函数，在此基础上精心构建了免干扰的 NHNCNN 能量函数。其次对NHNCNN的稳态判定进程加以改进以提高系统的收敛速度。特别地，采用类二分法联合NHNCNN去搜索最小信道分配总数。仿真结果表明，利用富足的NHNCNN时滞、噪声和混沌神经动力，所提算法能有效地搜索到合理解，并最终找到全局最优解，提高了频谱效率。与现有方法相比，所提算法能够实现更好的收敛速度和合理解率。
위료피면다매체다파조파단빈망(MBSFN)구역내부화구역지간적간우，진일보제고빈보효솔，해문제출일충개진적기우조성조절시체조성혼돈신경망락(NHNCNN)적동태신도분배방법。수선，근거MBSFN구역적특수탁복결구，중신정의료4충전자겸용한제함수，재차기출상정심구건료면간우적 NHNCNN 능량함수。기차대NHNCNN적은태판정진정가이개진이제고계통적수렴속도。특별지，채용류이분법연합NHNCNN거수색최소신도분배총수。방진결과표명，이용부족적NHNCNN시체、조성화혼돈신경동력，소제산법능유효지수색도합리해，병최종조도전국최우해，제고료빈보효솔。여현유방법상비，소제산법능구실현경호적수렴속도화합리해솔。
A dynamic channel allocation algorithm is proposed to avoid all interference and improve spectrum efficiency in Multimedia Broadcast multicast service Single Frequency Networks (MBSFN). Four electromagnetic compatibility constraint functions are redefined according to the topology information of MBSFN. In order to avoid all intra-area and inter-area interference of MBSFN, a novel energy function of Noise-tuning-based Hysteretic Noisy Chaotic Neural Network (NHNCNN) is constructed elaborately based on renewed constraint functions. Also, the judgment process of the stable state of NHNCNN is developed to accelerate system convergence. Specifically, the dichotomy method is adopted jointly to minimize the total number of allocated channels so as to further improve spectrum efficiency. Simulation results show that a feasible solution without any interference can be effectively searched by the improved NHNCNN. Finally, the optimal solution with minimum total channel number is found. Compared with existing algorithms, the proposed algorithm achieves better convergence speed and quality of solution.