Study On The Key Technology And Algorithms Of Broadband Beamforming

With the rapid development of society,the speed and capacity of the wireless transmission system are expected to be higher and higher,and the broadband signals are applied in communication systems widely.Antenna as the front end of the RF,is the primary channel to receive broadband signals,and there is an urgent need of broadband antenna array with a constant pattern for applications of the broadband communication system.Broadband beamforming is a technique that makes the pattern of antenna array approximate to the desired pattern in the broad frequency range by using beamforming technology,which can make antenna array receive he broadband signal with negligible distortions.Therefore,broadband beamforming shows great prospects in array signal processing field.Firstly,this paper summarizes the research status and overview of broadband beamforming technology at home and abroad.Then,the similarities and differences between broadband signals and narrowband signals are analyzed.Meanwhile,the problems of several common broadband beamforming methods are summarized.In addition,the principle and implementation flow of the particle swarm optimization(PSO)algorithm and the quantum particle swarm algorithm(QPSO)are analyzed.Secondly,this paper presents a novel broadband beam-former using a Bessel function based weight vector,aiming to solve the pattern distortion of wideband signals received by a uniform linear array.Owing to the similarity of the Bessel series and the pattern function of the uniform linear array,we design a weight vector based on Bessel functions of the first kind by deducing the relationship between the truncated position of Bessel series and the number of array elements.Simulation results show that the uniform linear array with a proper weight vector can receive wideband signal from 1 to 3 GHz with negligible distortions.The relative amplitude error for pattern functions at different frequencies remains below 10-3 and can be further reduced by increasing the number of the array elements.Thirdly,the PSO algorithm is studied and improved with oppositionbased learning.As known,the solution and the opposite solution both are closer to the global optimal solution with 50% probability higher than the other one.Therefore,the efficiency will be greatly improved when the solution and the opposite solution search at the same time.Simulation results show the PSO algorithm based on the opposition-based learning(OPSO)is superior to both the PSO algorithm and the QPSO algorithm in terms of convergent speed and precision.Finally,the OPSO algorithm is used to design the broadband beamformer of the circular array and uniform linear array.By adding the multiplication of the sub-narrowband pattern function and the desired pattern function,the method transforms the design of the broadband beam-former into a pattern synthesis problem with constant amplitude of the target pattern,and designs weight vector of circular array and uniform linear array with OPSO algorithm.Simulation results show that the circular array and uniform linear array with a proper weight vector can receive wideband signal from 1 to 3 GHz with negligible distortions.