基于熵测度优化的跳频信号参数估计

Parameter Estimation of Frequency Hopping Signal Using Entropy Measure Optimization

详细讨论了采用三阶Flandrin体积归一化熵测度随时频分布核参数的变化规律,给出了典型时频表示进行参数优化的方法,提出了一种基于熵测度的信号相关平滑伪魏格纳分布,对其在跳频信号参数估计中的应用进行了比较和性能分析.仿真结果表明了该量化测度在时频核函数性能评价时的可行性和有效性,基于熵测度优化的跳频信号参数估计有较好的抗噪声能力,可以取得比其他方法更精确的效果.

The evolution of third order Renyi entropy with normalization volume versus kernel parameter of time- frequency distribution was discussed. A method of parameter optimization for some time-frequency representations in terms of the criterion was given and the assessment of the distributions was accomplished by evaluating the measure. A novel signal-dependent smoothed pseudo Wigner-Ville distribution using Renyi entropy was proposed. Simulations resuits showed that this measure was feasible and efficient for evaluating the performance of different kernel, this signal dependent time-frequency analysis based on the measure was more robust and can yielded a more accurate estimation of frequency-hopping signal than other methods.