Research And Implementation Of Multi-Carrier Modulation System Based On Time-Varying Filter Bank

The forms of information transmission are gradually showing a trend of complexity and diversification.In the future,mobile communications will gradually develop toward intelligence and self-adaptation.The dynamic transformation of service types and the real-time configuration of physical resources pose unprecedented challenges to information transmission technology,which puts forward higher adaptability requirements for the coordination and cooperation of existing multiple alternative multicarrier modulation structures.This thesis mainly studies the modulation structure for future communication systems to meet the needs of flexible switching and adaptation of transmission waveforms.The main research content is as follows:1.In response to the business diversification of the future mobile communication system,this thesie proposes an adaptive orthogonal frequency division multiplexing(Adapt-OFDM)system based on discrete cosine transform.In order to improve the fast and flexible configuration ability of the modulation system,Adapt-OFDM configures the prototype function in sections,adjusts the number of subcarriers and the length of the filter to ensure the adaptive conversion of the modulation waveform during the user's transmission service switching process without interrupting the transmission.Through simulation and analysis of the spectrum change of the prototype function and the system bit error rate,it is concluded that Adapt-OFDM has good spectrum convergence characteristics and peak-to-average power ratio performance,which can achieve fast and flexible seamless switching,and can ensure that the perfect reonstruction of modulation system before and after information switching.Under the influence of the transmission channel,the bit error rate performance of Adapt-OFDM has no obvious loss compared with the OFDM system.2.In response to the future demand for intelligent mobile communications,this thesis proposes an intelligent carrier modulation structure design scheme combining timevarying filter bank structure and deep learning network.At the transmitting end,the solution uses a time-varying filter bank to adaptively adjust the number of subcarriers,prototype functions,and overlap coefficients and other important configuration parameters to change the transmission waveform.At the receiving end,a deep neural network is used to complete the recovery of the received waveform.Through simulation analysis,compared with the traditional orthogonal frequency division multiplexing(OFDM)system and filter bank multi-carrier modulation(FBMC)system,the solution has lower bit error rate and flexibility higher.The proposed solution not only solves the difficulties in the design of the receiving end of the traditional time-varying filter bank,but also provides new ideas for the research of new intelligent modulators.