AbstractComPared with a mechedtal scanning haar a phased mp radar has mp adVantages,such as sforultaneOus multiple beam and flekibly conbolled beam. Howevf, the haonalphased arra for can not caneel achve interference. With the teCboque of adaghve digitalbeaxn-fonnin rpBF), the ndhotal phed N radar could hav this ability For lnyeN with mny sensors, sensor level adaPtiv PrOCessing is comPUdsonally expenSbe andresultS in high coSt. In this case, the subW level adaPtive ProceSsing is a bethe tradcoffWn System --ce and cosL The Peak POwe of CW for is equal tO its averagePOwe and itS tIanSndtter can be easily realed with solid State COmPOenL Therefore, it hassmall volume, good mobility and can be tal on or offconVedentiy. In addihon. the CWM also has the proPerty of low probability of IntercePhon (LPD' Some key Problems onthe planar phased arra adaPtive DBF system for X band CW the are reasehed in thesdissertation. the main work is sununarized as follows:For the adaPtve DBF ler the anenn System is closely related ed signal ProCessingtecndque and is an driPOrtan faCor in the System cost. In chaPter l, tWo kinds of low coStphased W anenna systeIn. namely sparse mp and mulh-bo pe fed axray areanalyzed and comPared. A principle for SUbanay division and a differenee beaIn Synthesismethod at submp level for SParse anay are ProPOsed. The authr conSders that the multi-feed sPace feed W is more sultable for X band for Under mpnt condion.It is well known tha the mp ermr may grealy degrade the perfOnnanee of adaPtiveN signal Processing, so the arra error calibration is very twrtan for adaPve DBFler In thes dissertaion, by means of a edbndion source and mp senSOs PhaSe Shifterwe proPOsed a algorithIn for joint1y calibhan the sensors amPitude and PhaSe ermr, sPacetwt parameter error and Phase shifter error for space fed anay The algorithIn alsoholds in other anas with subarra de.The adaPive DBF aIgorithIn for CW ed is qulte differen from that for the Pulse edAner analyzing the featUr of the CW radar we advanced a SPatial blocking fiiter derivativeconSwt algorithm for planar W based on Fudge's 1inear anay algorithIn, and comPareditS pefformance with derivative conStrained LCMV algorithm. The shaulthen boindcate tha the proPOsed algorithIn is bettr than the derivthee constrained LCMValgorithIn because it avoids the signal caneellation and adaPtive Inain beam distonioneffectively without the gain loss and broadening main beam.The azimuth and element angle are coupled in the monopulse angle estimation for the planar array. We proposed a derivative constrained beamforming algorithm for separating the two-dimension angle information and simplifying the 2-D angle estimation algorithm to l-D angle estimation algorithm. Moreover, it can result higher angle estimation precision than the first order approximation algorithm.In practical applications, the relative motion between radar platform and interference can broaden the direction angle of interference signals and severely degrade the performance of adaptive array if the weights adjustment is not fast enough. In chapter 6, we advanced an adaptive beainforming algorithm with GSLC structure and robustness against janlmer motion. The simulation results indicate that the proposed algorithm performs better than the conventional GSLC algorithm in moving interference scenarios.Adaptive blind beamforming algorithm is also an effective method for extracting the required signal in the existence of array errors. We proposed a blind adaptive beamforming algorithm based on the signal抯 self-coherence property. Employing the estimated array manifold, the required signal can be extracted while eliminating the interference. We also discuss the application of the algorithm to CW radar.Besides sensor amplitude and phase error, mismatch between channels can also degrade the performance of adaptive arrays. In chapter 8, we proposed an adaptive blind equalization algorithm based on the cyclostationaiy pro