一种高速跳频信号的多载波频点提取方法OA

In view of the multi-user separation challenge in non-cooperative high-speed frequency-hopping signals in non-cooperative communication scenarios,this paper proposes a rapid detection method based on mutation characteristics of time-frequency domain power vector.By constructing the dynamic difference operator for the local maximum frequency power vectors,a hop timing estimation model within a time-frequency analysis framework is established to realize the joint extraction of multi-user frequency parameters.The proposed method adopts a first-order differencing strategy on time-frequency power spectrograms,significantly reducing the computational complexity compared to conventional energy accumulation algorithms.Experiments show that when the equivalent signal-to-noise ratio is greater than or equal to 12 dB,the proposed method achieves a user detection probability of 98.2%,and its processing latency is reduced by 56%in comparison with the traditional sliding window methods.To sum up,the proposed method provides a solution with engineering practical value for blind separation system of frequency-hopping signals by hardware-friendly computational architecture.