| The filter bank multi-carrier with offset quadrature amplitudem(FBMC/OQAM)system has become a powerful alternative to frequency division multiplexing(OFDM)system and one of the key technologies of the fifth-generation(5G),which has the advantages of lower side lobe energy,more robust asynchronous transmission performance,and higher spectrum utilization.In this thesis,we will shift our focus to two performance-determining technologies in FBMC/OQAM system.The system advantages of FBMC/OQAM depend on the design of a prototype filter with high time-frequency focus,which means low side lobe energy.However,the 5G communication scenarios are complex and diverse,and none of the current design methods can perform oriented side lobe energy suppression according to the needs of the communication scenarios.On the issue of the prototype filter design for oriented side lobe energy suppression in FBMC/OQAM,the inter-carrier and inter-symbol interference formulas when the FBMC/OQAM signal is under imperfect reconstruction conditions are derived according to the FBMC/OQAM system model.Then the dual-objective optimization problems are constructed and the corresponding solving algorithm is proposed later.The dual-objective optimization problem minimizes the energy of the stop band with the constraint factor,and the comprehensive interference between the channel and the symbol is controlled under a certain threshold range.The prototype filter coefficients also should be symmetrically distributed and have Nyquist properties.Considering the uncertainty of the constraint factor in the objective function,a sub-optimization problem is constructed,which minimizes the side lobe energy of the second or third segment by limiting the side lobe energy and total stop band energy of the first segment to a certain threshold range.Then,in order to solve this dual-objective optimization problem,we propose the nested sequence quadratic programming genetic algorithm(NSGA).The algorithm uses Genetic Algorithm(GA)to obtain the optimal constraint factors,and then uses Sequential Quadratic Programming(SQP)to obtain the optimal prototype filter coefficients.Finally,the numerical results verify that the prototype filter designed in this paper has lower stop band energy than the existing prototype filter with good performance,and can achieve oriented side lobe energy suppression within specified segments.Moreover,the algorithm proposed is verified its convergence characteristics under simulations.In addition,the FBMC/OQAM system has an inherent disadvantage similar to the OFDM system that is Peak-to-Average Power Ratio(PAPR).The high PAPR will affect the distortion and data transmission rate of the FBMC/OQAM signal.The method based on the partial transmission sequence(PTS)technology performs well in the reduction of PAPR,but as the number of sub-block division increases,the computational complexity will also increase.On the issue of how to use PTS technology to reduce the PAPR of FBMC/OQAM signals and reduce the computational complexity,we propose a GA-based bilayer partial transmission sequence(GA-BPTS)scheme,which considers the difference in signal structure between the FBMC/OQAM and the OFDM.The GA-BPTS scheme continues to divide each sub-block,converts the single-layer structure of the conventional PTS into a bilayer phase factor search structure,and introduces a penalty operation.Then,a genetic algorithm is used to obtain the suboptimal phase factor vector.In particular,with the help of the prototype filter design principles in the previous chapter,before the PAPR reduction process,the GA-BPTS scheme uses a new prototype filter that has a better out-of-band attenuation performance which is more effective to reduce the PAPR of the FBMC/OQAM signal.Simulation has confirmed that,compared with the traditional PTS scheme,the GA-BPTS scheme not only provides better PAPR performance,but also reduces the computational complexity for FBMC/OQAM.There are 28 figures,8 tables and 90 references in this paper. |
PDF Full Text Request