Research On Waveform Technology Of Wireless Network Cognitive Communication Signal Based On Artificial Intelligence

With the development of communication measurement and control technology and the rise of new artificial intelligence technology,intelligence and informatization become the development direction of wireless communication in the future.But the spectrum resources are increasingly scarce,the communication environment is complex and changeable,and the signal will be interfered by all kinds of clutter in the transmission process.The proposal of cognitive dynamic system(CDS)brings dawn for wireless communication technology.Its adaptive waveform technology can effectively avoid interference and improve the system anti-interference ability and target detection performance when the target characteristics or external environment changes.Adaptive waveform technology can improve the performance of cognitive communication by sensing the external environment information and adjusting the relevant parameters.In the existing theory and literature,only its application in cognitive radar target detection and tracking is studied,lacking of other innovative scene combination.Neural network algorithm can realize the fitting of nonlinear information system through a large number of structured data mining rules.Based on this,combined with the measurement and control communication scene,the adaptive waveform system is constructed based on the minimum bit error rate criterion,and the BP(Back Propagation)neural network is used to construct the BP network adaptive parameter learning model,which makes the cognitive communication more efficient and intelligent.The research work and innovation of this paper are as follows:(1)Combined with measurement and control communication scene,taking the SER(Symbol Error Rate)as the objective function,and based on the minimum SER criterion,an adaptive optimization strategy under various interference environments is studied.Based on the theory of cognitive waveform technology,the adaptive optimization models of waveform parameters and environment parameters are constructed by using LFM(Linear Frequency Modulation)signal and combining four kinds of noise environments,including White Gaussian Noise(WGN),Lognormal Noise(LN),Weibull Noise(WN)and Single-Tone Noise(SN).The simulation results show that the SER performance of the system is related to the selection of transmit waveform parameters and environmental parameters.Combined with a variety of noise environment models,reasonable adjustment of transmit waveform parameters and environmental parameters can effectively improve the cognitive transmission performance and communication quality.(2)Combined with BP neural network algorithm,the optimization model of BP network cognitive waveform parameters is constructed.The sample data set is obtained by adaptive waveform system,the pulse width T and SNR(Signal to Noise Ratio)are used as input training data,and the SER(Symbol Error Rate)is used as output training data.The adaptive parameter learning strategy of BP network is proposed.The prediction performance and prediction error are analyzed by matlab code simulation,and the influence of sample number on prediction performance is studied.The simulation results show that the BP algorithm model proposed in this paper has good prediction effect.Compared with the adaptive waveform algorithm,the SER prediction curve optimized by BP algorithm has better convergence effect,closer to the theoretical value,and effectively reduces the computational complexity of code implementation.