Research On Waveform Design For Adaptive Radar In Spectrum Congested Environment

Radar obtains the position and motion information of the target by transmitting and receiving electromagenic signal.In order to get a sufficient range resilution,conventional radars commonly need to occupy a continuous wide band.However,with the growing number of electromagnetic users and their increasing demand for spectrum bandwidth,the radar will inevitably share the identical frequency band with other electromagnetic users,increasing the likelihood of mutual-interference between radar and other users.To ensure the spectral compatibility in a spectrum-congested environment,radar system needs to avoid those frequency points/segments in which exisit strong interference or users of higher priority,which however inevitably degrades the autocorrelation performance of the transmit waveform.To address these problems,modern radar systems need to better "cognize" the working environment and fully exploit various degrees of freedom of the waveform to care its own performances.To achieve this goal,this paper analyzes the performance of radar transmit waveforms in spectrum congested environment,gives the algorithm for parameters optimization of the waveform,and realizes the optimal design of several different forms of waveforms.In the analysis of waveform performance,this paper establishes the system for evaluating the range resolution and range sidelobe level of waveform under given working frequency band and provided a method to give rapid predition on these two metrics.In this work,we describe these two metrics both as the function of the power spectrum of the waveform and establish the corresponding optimization problem.We derive the analytical expression of the optimal power spectrum using the Lagrangian method and then give prediction on the optimal performance.With this work,the radar system can better "interpret" the external spectrum environment and thus work more intelligently.In optimizing the waveform parameters,this paper analyzes the commonness of optimizations of different metrics and unifies them as spectrum optimization.We utilize the residual error between the waveofm?s spectrum and a prescribed spectrum template as the unified performance metric and obtain waveform with different desired features by minimizing the unified metric using congugate gradient method.This paper devises a way to employ FFT in calculating the above menthioned spectrum residuals and its gradient,greatly reduing the computational complexity for minimizing this metric.Using the scalar method and the chain rule,the proposed method can also flexibly solve those complex optimization problems in which multiple design goals are to be considered simultaneously.The above work lays the algorithm foundation for the optimal design of different forms of waveforms in this paper.In designing specific waveforms,this paper establishes the schemes for optimizing PCFM(Polyphase-Code Frequency Modulation)waveform,interpulse diversed waveform,and frequency agile waveform in spectrum congested environment.PCFM waveform is a kind of intrapulse fast-time modulated waveform proposed in recent years.Due to the constant amplitude envelope and the phase modulation function with continuous transitions,the PCFM waveform can maximize the average transmit power of radar and effectively control its spectral sidelobes.To address the problem that the existing schemes for PCFM waveform design are mostly complex and time consuming,this paper proposes a design scheme based on interpolation and reconstruction,which transforms the optimization of PCFM waveform into the optimal design problem of its uniformly sampled sequence.Further,this paper describes the optimization of the sampled sequence as a fitting problem of its power spectrum,and subsequently achieves rapid solution using the spectrum optimization tool proposed in this paper.Based on the performance analysis of radar waveform in spectrum congested environment,we can realize a comprehensive control on the range resolution and range sidelobe level by selecting the appropriate target spectrum template.The experimental results show that the designed PCFM waveform can effectively suppress the mutual interference with other in-band users and meanwhile ensure the autocorrelation performance,while having a very "clean" spectral sidelobeThe interpulse diversed waveform refers to a waveform whose transmitted pulses are changing during a signal accumulation period.Due to the higher degree of freedom,the interpulse diversed waveform has stronger anti-jamming capability than the coherent pulse train.Moreover,when significant changes of the external spectral environment occur in an accumulation period,the interpulse diversed waveform would be the inevitable choise of an adaptive radar system.In view of the above facts,this paper establishes the design model of adaptive interpulse diversed waveform in time-varying spectral environment and proposes the design scheme.In waveform design,this paper considers two cases in which the transmit pulses are respectively phase-coded signal and PCFM waveform,and proposes the method to optimize the signal spectral compatibility and autocorrelation simultaneously.Unlike most of the existing studies,this paper takes into account the characteristics of already transmitted pulses when optimizing the autocorrelation performance,rather than solely optimizing the pulses to be trainmited.Due to the global perspective utilized,the proposed design can achieve improved overall performance than the traditional scheme in a time-varying spectral environment.Frequency agile waveform is a kind of interpulse frequency diversity signal.It has the advantage that larger range resolution bandwidth can be obtained without the hardware demand for increasing system instantaneous bandwidth and sampling rate,compared with intrapulse fast time modulated waveform.In addition,by properly allocating the transmit frequencies,the radar system can lower the range sidelobe level to the most extent while avoiding mutual interference.However,as the optimization problem of frequency agile waveform is highly nonconvex,how to rapidly obtain the global optimal frequency combination has always been a very difficult problem.To address this problem,this paper achieves the fast and optimal design of the transmit frequencies in two steps.In the first step,this paper obtains an amplitude function with the desired autocorrelation characteristic via solving a convex optimization with polynomial complexity and then allocates the transmit frequencies with different density to simulate the amplitude function to obtain an approximated estimation of the optimal solution of the original problem.In the second step,this paper employs the gradient method to locally optimize the approximated solution obtained and finally obtains the(qusi-)optimal solution.Experiments show that the proposed design method can achieve the optimal design of the frequency agile signal with very little computation time.