Research On Non-adaptive Clutter Suppression Methods And Application Forairborne Radar

Motion features of airborne radar will make the clutter exhibit strong space-time coupling characteristics,in order to realize effective suppression of airborne radar clutter, STAP(Space Time Adaptive Processing) has been developing unprecedentedly in recent decades. Generally speaking, clutter is uniformly distributed in conventional STAP, but actually the real environment of airborne radar is quite complex, which makes conventional STAP difficult to achieve effective result of clutter suppression,the development of radar techlonogy brings the rapid growth of antenna elements and high dimension of signal processing,which raises a higher requirement for the performance of radar's real-time processing. Aiming at high dimension and small scale of airborne training samples, it is extremely urgent that we study airborne radar's signal processing technology aiming at high dimension and small scale of training samples. This thesis introduces non-adaptive clutter suppression methods of low computational complexity and small scale of training samples on the background of airborne radar and mainly consists of the following content.1. Chapter 2 mainly introduces the Non-Adaptive Two Dimension Pulse-to-pulse Cancellation(Non-ATDPC) algorithm,which is on the background of airborne radar and realized by constructing a cancellation coefficient matrix within two adjacent pulses during a Coherent Processing Interval(CPI). The cancellation coefficient matrix is only influenced by airborne radar's operating parametres and geometric information,most clutter energy can be suppressed by cancellation processing.The advantage of Non-ATDPC algorithm is not dependent on large scale of training samples and avoids a number of operations. The crab of airborne radar is also considered during the design process of this algorithm,so Non-ATDPC can be used in various states of airborne radar(side-looking, strabismus-looking,front-looking) and therefore effectively broaden its practical range, Non-ATDPC algorithm can be seen as the pre-filtering processing before classical dimension-reduction algorithms such as Factored Approach(FA),Extended Factored Approach(EFA),etc. Subsequent simulations also confirm that Non-ATDPC has a better clutter-suppression performance compared with simply executing dimension reduction operations such as FA,EFA,etc.2. Chapter 3 mainly introduces Space Two-Stage Dimension Reduction(S-TSDR) method,aiming at real problems airborne radar encounters in real environment-small scale of training samples and a large number of array elements,conventional FA?EFA algorithm which is realized by firstly domain doppler processing and then spatial adaptive processing has the problem of insufficient sample demand,which results in the degradation of radar's clutter suppression.Aiming at the problems mentioned above, the S-TSDR algorithm we propose can reduce the requirement for training samples and operation amount.The main idea of S-TSDR is dividing the long spatial vector obtained after domain doppler processing into two short vectors using Kronecker properties,so we can get a couple of quadratic cost functions and get its optimal solution through the use of loop iterations.Through rapid convergence of this algorithm we can effectively improve the real-time performance of airborne radar's signal processing.3. Chapter 4 mainly introduces Non-Homogeneous Clutter Two-Dimension Pulse-to-pulse Cancellation(NHC-TDPC) algorithm on the basis of city building model, which is the practical application of Non-ATDPC mentioned in Chapter 2. This Chapter firstly constructs a city building model and then establishes the non-homogeneous clutter model on the basis of that. Conventional STAP methods have a bad performance on the suppression of non-homogeneous clutter, threfore we propose the NHC-TDPC algorithm.The core idea of NHC-TDPC is to construct a pulse cancellation coefficient matrix which is only connected with airborne radar's operating parameters and spatial geometric information regardless of radar's receiving sample data, we can process the training sample data using the obtained pulse cancellation coefficient matrix,we can get both the true clutter echo signal from the range circle to be detected and estimated clutter echo signal from all effective range circles except one which is to be detected. We can adjust the distance between the airborne radar and the city building model so that airborne radar can exhibit different states(short-range, medium-range and remote-range).Both the power spectrum and characteristic spectrum can be obtained corresponding to various states of airborne radar, by analyzing and comparing the power spectrum and characteristic spectrum obtained we can selectively put the NHC-TDPC algorithm into effective use.