The recently proposed coprime array, as an example of a sparse array of antennas, could produce more effective virtual sensors (degrees of freedom (DOFs)) than the number of actual sensors being used. The coprime array achieved that by utilizing the coprimality property of two coprime integers M and N to build its two consisting uniform linear subarrays, M-array and N-array, respectively. The N-array is of N sensors with an inter-sensor spacing of units, while the M-array is of 2M actual sensors with an inter-sensor spacing of units.
　　Inthisthesis,adirectionofarrival(DOA)estimationbasedontheoptimizedcoprimearray(OpCA)configurationisproposed.TheproposedarraykeepstheN-arrayuntouched,whileitshiftsallthesensorsoftheM-arraybyaspecificdisplacementamountLstowardsthepositivedirection,yieldingtheOptimizedM-array.DifferentdisplacementexpressionsforLsareproposedforthepurposeofbeingapplicabletoallthepossibleencounteredcoprimeMandNvalues.CoprimeMandNintegers,whereM＜N,canbeclassifiedintofivepossibletypesandaproperdisplacementexpressionforeachisassignedaccordingly.Analytically,thosedisplacementsareoptimizedsuchthatthenumberoftheactualsensorsfortheproposedarrayisstillthesameasthatforthecoprimearray,(N+2M-1),viahavingacommonsensorsharedbetweenthetwosubarrays,andthenumberofcentralcontiguouslagsorequivalentlyvirtualsensors,whichreferredtoastheuniformdegreesoffreedom(DOFs)capacityinthedifferencecoarray,hasbeenenhanced.Thisenhancementisshownbyincreasingtheone-sideuniformDOFscapacityofthecoprimearrayfromMN+M-1toMN+M-1+Ls+MwhereLsisthedisplacementvalueitself.Thisenhancementalsocomeswithreducingthenumberofactualsensorpairsofsmallseparations(smallmultiplesofλ/2),therebybringinglessmutualcouplingcomparedtothecoprimearray.Numericalsimulationresultsusingsubspace-basedDOAestimationalgorithmsareprovidedtodemonstratetheeffectivenessoftheproposedgeometry.