Blind Adaptive Interference Suppression For Civil Aviation Air-ground Communication

With the rapid development of the civil aviation industry, air traffic flow is increasingcontinuously, and the requirements for the control means of the air traffic control departmentsare becoming higher. At the same time, the continuous growth of the national economy haspromoted the expansion of the telecommunications business, and the electromagneticenvironment became worse and worse. Due to the various communication equipments on themarket, and the weak market control, some inferior products enter into the market, and theproblem of electromagnetic interference is more and more serious. Consequently theair-ground communication distance is shortened, and the noise and speech distortion makecommunication between pilots and air traffic controllers even difficult, which poses apotential threat to flight safety.At present, strengthening the radio management or the radio check by the expensiveprivate check vehicles and aircrafts are the main means to resolve the interference problem,but they are just temporary solution. This dissertation proposes that the blind signalprocessing techniques can be used to solve the interference problem in air-groundcommunications of the civil aviation. According to the properties of the interference signaland different application, the single-channel constant modulus interference suppressionscheme (for suppressing the constant modulus interference signal, applicable to airbornestation), the bi-channel constant modulus interference suppression scheme (for suppressingthe constant modulus interference signal, applicable to ground station), and the bi-channelnon-constant modulus interference suppression scheme (for suppressing any form interferencesignal, applicable to ground station) are adopted. The algorithms for the interferencesuppression are investigated and the results are as follows:1. Taking account of the constant modulus characteristic of most interference signals inair-to-ground communication of civil aviation, single-channel optimal constant modulusalgorithm is proposed and applied to the interference suppression scheme of signal-channelconstant modulus. In this algorithm the amplitude and phase of the constant signal areestimated by nonlinear least squares method, batch and recursive realize methods are given,and the convergence of the recursive manner is proved. Cooperating with frequencyestimation and notch filter, the interference signal could be suppressed effectively. The newmethod has some advantages, such as it could avoid the convergence problem of the commonadaptive interference suppression (local convergence, the selection of the initial weight andthe step-size), and it could be compatible with the existed communication systems, and thenumber of channel is not increased. Simulation and experimental results demonstrate that the new method could estimate the constant modulus signal effectively, and the quality of thevoice signal output by interference suppression system has been significantly improved.2. For solving the convergence problem of the LMS constant modulus algorithm in theconstant modulus array, the initializing method is proposed based on the LS constant modulusalgorithm. In this algorithm, small snapshot signal is used to provide initial weight vector forLMS constant modulus algorithm by LS constant modulus algorithm, and the weight vector isupdated by the LMS constant modulus algorithm. The new method has good interferencesuppression performance, and the convergence speed is fast and has no relation to theinterference environment, on the other hand, the extra computational burden is little. Thismethod is used in the bi-channel constant modulus interference scheme, simulation results onsimulation data and real data show that the new method improves the convergence speed ofthe constant modulus algorithm, and enhances the effect of the interference suppression.3. In order to improve the convergence speed of the constant modulus algorithm and toreduce the steady-state offset, the normalized variable step-size constant modulus algorithmbased on error is proposed. The error function in the update equation of the constant modulusalgorithm is introduced, and the step-size is adjusted by it. This method not only overcomesthe effect of the distribution of the data correlation matrix eigenvalue for the constantmodulus algorithm convergence speed, but also is consistent with the idea of variablestep-size.4. The new blind signal extraction algorithm is proposed based on the circularity ofcomplex signal, and it reduces the high computation burden for the interference suppressionscheme based on blind signal extraction. The circularity and second order independence of thewhitened data are used adequately, so the cost function is simplified. At the same time, theimproved recursive least square adaptive canceller is used to reform the canceling effectively.Simulations and experiments based on simulation data and real data verify the validity of themethod.5. Multi-variable step-size EASI algorithm is proposed, and it is applied to theinterference suppression scheme based on blind source separation. According to theorthogonality part and independence part in the iterative equation, the nonlinear functionsbetween the step-size of the orthogonality and independence are established, which are usedto adjust the step-size. The new method could track the change of the channel, improve theconvergence of the EASI algorithm, and enhance the effectiveness of interference suppression.An effective initializing method for EASI algorithm based on FastICA has been developed. Inthe new method, small snapshot FastICA algorithm is used to calculate the weight vector, and the conjugate transpose of the weight vector is multiplied with whitening matrix, and themultiplying result is provided to EASI algorithm as the initial weight vector，then uses EASIalgorithm to update the weight vector. The new method could enhance the convergence speedof the EASI, and avoid the convergence problem of the random choice of the initial weightvector. The simulation results based on simulation data and real data prove the fastconvergence and the performance of the new method in interference suppression.