Research On Radar Waveform Design Based On Information Theory

Cognitive radar can sense and learn the ambient environment. With the knowledge obtained from the environment, it can adjust the transmitted waveform by closed-loop feedback system and detect the target in different environment adaptively. It can overcome the disadvantages of the single transmitted signal of traditional radar and the difficulty to adapt to the complex working environment; hence it has become a popular research in recent years.Waveform design is one of the key problems of cognitive radar research. The designed radar waveform based on information theory can maximize the mutual information between random target impulse response and the received radar waveform, so it can get more target information. Therefore, researching on radar waveform design based on information theory has the very important theoretic value and practical significance. This paper concluded the adaptive waveform design method of cognitive radar and researched the maximum marginal allocation algorithm and constant modulus algorithm.Due to the clutter and noise exist at the same time in practical application, we introduce clutter into the formula of mutual information and give the detailed formula derivation of the relationship of mutual information and transmitted waveform in clutter and noise environment. Although the use of a Lagrangian multiplier approach will produce the global maximum, the search of the multiplier will become complicated in some situations. In order to solve this problem, we introduce the maximum marginal allocation algorithm to radar waveform design. Simulation results show that using this algorithm can also get the optimum radar waveform and the algorithm is very simple and computationally efficient. This algorithm has a very good universal compared with waterfilling algorithm and shows a new way to solve the optimal radar waveform.Modern radar usually designs transmitted waveform according to single task. Aiming at this problem, a design framework combining both the target detection and estimation performance is presented in clutter environment, and then we apply the constant modulus algorithm to radar waveform considering the launching process of the transmitted signal. Simulation results show that the obtained waveform can fulfill the task of target detection and estimation in slow time-varying conditions. After applying the constant modulus algorithm to the obtained signal, it has constant envelope property in time domain and has the given Fourier transform magnitude, and can also complete the target detection and estimation task.