Research And Implementation Of Key Technologies For Link Adaptive Decision Based On Interference Sensing

The perception of human interference in the electromagnetic environment and the response to the changing electromagnetic environment through link adaptive decision techniques are becoming increasingly important.The thesis investigates interference sensing techniques,uses the sensed interference information as input for link adaptive switching,designs the overall interference sensing link and data transmission link,and completes the FPGA implementation of the downlink data transmission link.The thesis selects an energy-based detection algorithm for a variety of typical interferences such as pulse interference,single-tone interference,multi-tone interference,partial-band interference and linear-sweep interference,and completes the interference detection by implementing an adaptive threshold through the FCME algorithm.An interference detection scheme with time and frequency domain cascades is designed according to the different characteristics of interference in the time and frequency domains,and the detection rate of all five types of interference can reach more than 0.9at a JNR of 0d B.Subsequently,the study of signal,interference and noise power estimation is carried out.Then,this paper carries out the study of signal,interference and noise power estimation,using the synchronous sequence to estimate the channel fading and eliminate the influence of the signal sequence to obtain the power of the signal;then the average power of the noise and the interference power are estimated in segments,and finally the Signal-to-Noise Ratio(SNR)and the interference power are estimated.The numerical simulation shows that the NMSE of the SNR estimation under different interferences is less than-15 d B when the SNR is 0d B,and the NMSE of the interference power estimation under any interference can be less than-15 d B.Only pulse interference can be detected in the time domain,so the detection in the time domain is the identification of pulse interference.For the other four interferences,the thesis compares and analyses a variety of typical features based on the performance of interference in different domains,which are frequency domain moment kurtosis coefficient,normalized spectral kurtosis coefficient,normalized spectral 3d B bandwidth and normalized spectral long-short signal bandwidth difference.Based on the four features,the interference classification schemes based on decision trees,SVM and BP neural networks were designed,respectively,and numerical simulations showed that the overall classification rate of single machine learning methods could not reach 0.9 at 0d B.Therefore,by combining decision trees with SVM and chunking a fourfold classification into three binary classifications,a decision tree-SVM cascade interference detection scheme is designed,and numerical simulations show that the scheme can achieve more than 0.9 interference detection rate at 0d B.The thesis further investigates the interference detection scheme with CNN as the classification model,which can automatically abstract high-level features from the original data and complete the detection by adding pure noise to the training set.Simulation results show that both the four-classified and sixclassified CNNs can achieve more than 0.9 interference detection accuracy at 0d B.When compared with the traditional scheme of detection followed by classification,the CNNbased six-classification interference classification scheme under the current JNR has similar performance to the traditional scheme,but higher complexity,so the subsequent overall link uses the traditional scheme for interference detection and classification.In order to obtain as much transmission efficiency as possible while ensuring communication reliability,Thesis investigates the link adaptive decision technique with interference class,interference power,SNR,and interference bandwidth as input and optimized link as output.Thesis investigates the link adaptive decision technique based on ABC,which can perform local and global search for optimal links in the current electromagnetic environment by optimizing the fitness function.Through simulation comparison,the link adaptive technique based on the bee colony algorithm can obtain higher transmission efficiency than the traditional table look-up method under the BER threshold.Thesis constructs an overall data transmission link using link adaptive technology.The uplink senses the current electromagnetic environment during the silent period,and then decides the optimal link based on the information obtained from the sensing and feeds it to the downlink.The downlink changes the transmission mode according to the optimal link suggestion to ensure the reliability and efficiency of communication.Thesis designs the key modules of the downlink based on FPGA,including CRC module,Code-Rate Matching-Modulation module,Interleaving module,IFFT module,synchronization module,and channel estimation and equalization module.