SYSTEMS FOR THE GENERATION OF COMPLEX WAVEFORMS IN NEAR-FIELD FOR ANTENNAS AND RADAR SYSTEMS CHARACTERIZATION

THIS THESIS ADDRESSES THE DESIGN, IMPLEMENTATION, AND PERFORMANCE ASSESSMENT OF COMPLEX WAVEFORM GENERATORS (CWGS), NAMELY SYSTEMS CAPABLE OF RADIATING COMPLEX ELECTROMAGNETIC WAVEFORMS WITHIN PRESCRIBED NEAR-FIELD (NF) REGIONS, REFERRED TO AS QUIET-ZONES. THE RESEARCH ACTIVITY COMBINES TWO COMPLEMENTARY ASPECTS: THE DEVELOPMENT OF NOVEL TECHNIQUES DEVOTED TO THE SYNTHESIS OF SUCH SYSTEMS AND THE DEVELOPMENT OF INNOVATIVE STRATEGIES REQUIRED TO VALIDATE THEIR PERFORMANCE AND ENSURE THEIR ACCURATE CHARACTERIZATION. WITHIN THE SYNTHESIS’ FRAMEWORK, A NOVEL TWO-STEP APPROACH IS DEVELOPED FOR THE DISCRETIZATION OF EQUIVALENT RADIATING PANELS, PROVIDING A PRACTICAL IMPLEMENTATION THROUGH NONUNIFORM ARRAYS. IN THE FIRST STAGE, A GENERALIZED GAUSSIAN QUADRATURE RULE IS EXPLOITED TO DISCRETIZE THE RADIATION INTEGRALS, PROVIDING A SET OF NODES, NATURALLY RETAINED AS THE ARRAY ELEMENTS’ POSITIONS, TOGETHER WITH A CORRESPONDING SET OF WEIGHTS. THE RESULTING DISCRETIZATION SERVES AS THE STARTING POINT FOR A CONSTRAINED OPTIMIZATION PROCEDURE, WHERE THE ARRAY GEOMETRY IS OPTIMIZED UNDER REALISTIC DESIGN CONSTRAINTS, ENSURING PRACTICAL REALIZABILITY. IN THE FINAL STAGE, ONCE THE OPTIMIZED GEOMETRY IS FIXED, THE EXCITATIONS REQUIRED TO SYNTHESIZE THE DESIRED NF WAVEFORMS ARE DETERMINED BY LEVERAGING A SINGULAR VALUE DECOMPOSITION APPROACH. THE METHODOLOGY IS INITIALLY DEVELOPED FOR THE DISCRETIZATION OF A 1-D PANEL AND SUBSEQUENTLY EXTENDED TO THE CORRESPONDING 2-D CASE. NUMERICAL RESULTS ASSESS THE EFFECTIVENESS OF THE APPROACH IN SUCCESSFULLY GENERATING COMPLEX NF WAVEFORMS. THE SECOND PART OF THE THESIS ADDRESSES THE PROBLEM OF ACCURATELY CHARACTERIZING SUCH SYSTEMS BY DEVELOPING DEDICATED TECHNIQUES FOR THE CORRECTION OF 3-D PROBE POSITIONING ERRORS AFFECTING NONREDUNDANT (NR) NF MEASUREMENTS. TO THIS END, A DEDICATED TWO-STEP CORRECTION STRATEGY IS INTRODUCED. IN THE FIRST STEP, A PHASE CORRECTION TECHNIQUE COMPENSATES FOR THE SHIFTS OF THE SAMPLES FROM THE NOMINAL SCANNING SURFACE, AND THE CORRECTION PARTICULARIZES DEPENDING ON THE ADOPTED SCAN GEOMETRY. IN THE SECOND STEP, AN ITERATIVE ALGORITHM COMPENSATES FOR THE RESIDUAL 2-D ERRORS AFFECTING THE NF DATA BY EMPLOYING AN AD HOC ITERATIVE SCHEME. IN THIS THESIS, HELICOIDAL AND SPHERICAL SCANS ARE EXPLICITLY CONSIDERED. THE VALIDITY OF THE PROPOSED APPROACHES IN COMPENSATING EVEN FOR SEVERE 3-D POSITIONING ERRORS AFFECTING THE NF DATA ARE ASSESSED NUMERICALLY AND EXPERIMENTALLY. FINALLY, NR SAMPLING STRATEGIES TO REDUCE THE NUMBER OF REQUIRED NF SAMPLES IN SPHERICAL SCANNING CONFIGURATIONS INVOLVING OFFSET-MOUNTED ANTENNAS ARE PRESENTED AND EXPERIMENTALLY VALIDATED. THE RESULTS DEMONSTRATE THAT THE AMOUNT OF NF DATA REQUIRED BY EMPLOYING THIS TECHNIQUE COINCIDES WITH THAT REQUIRED IN THE ONSET-MOUNTING CASE, CONFIRMING THE EFFICIENCY AND PRACTICAL RELEVANCE OF THE PROPOSED TECHNIQUES IN REAL MEASUREMENT SCENARIOS.