ONLINE, QUASI-STATIONARY, CHANNEL IMPULSE RESONSE MODEL FOR NUCLEAR DISTURBED RF PROPAGATION

WE PROPOSE TO INVESTIGATE THE FEASIBILITY OF AN ONLINE, CHANNEL IMPULSE RESPONSE MODEL FOR RF PROPAGATION THROUGH NUCLEAR-DISTRUBED IONOSHPERES. THE PROPOSED MODEL WOULD GENERATE REALIZATIONS OF THE RECEIVED SIGNALS AT A SET OF ANTENNAS BY LINEARLY COMBINING INDEPENDENT RANDOM VARIABLE. THE CORRECT LINEAR COMBINATION WOULD BE DETERMINED FROM THE MATRIX OF CORRELATION COEFFICIENTS FOR THE VOLTAGES AT THE SET OF RECEIVING ANTENNAS. FOR STATIONARY CHANNEL CONDITIONS AND FIXED ANTENNA ORENTIATIONS THIS METHOD IS VERY FAST AND REQUIRES ONLY A SMALL ARRAY SPACE. FOR NON-STATIONARY CONDITIONS THE LINEAR COEFFICIENTS CAN BE PERIODICALLY UPDATED THEREBY ALLOWING THE SIGNAL REALIZATION TO EVOLVE IN A QUASI-STATIONARY MANNER. THIS MODEL IS KNOWN TO BE FEASIBLE UNDER RESTRICTED CHANNEL CONDITIONS (THE TURBULENT MODEL). WE WOULD EVALUATE WHETHER THE MODEL CAN BE EXTENDED TO ENCOMPASS THE DNA GENERAL CHANNEL IMPUSLE RESPONSE MODEL THAT IS CURRENTLY IMPLEMENTED IN ACIRF/RCIRF. THE ADVANTAGES OF THE PROPOSED CHANNEL MODEL, RELATIVE TO CURRENT MODELS, ARE: 1) SIGNAL REALIZATIONS PROVIDED DIRECTLY TO A COMMUNICATIONS/RADAR SIMULATION BY MEANS OF A SUBROUTINE CALL; 2) CHANNEL CONDITIONS AND ANTENNA ORIENTATION COULD CHANGE CONTINOUSLY WITH TIME; 3) SMALLER RAM, DISK STRAGE AND CPU REQUIREMENTS THAN CURRENT METHODS.