The use of diversely polarized antennas in an array system can significantly improve the overall's array performance. However, its capability does not get utilized to its full potential when the conventional array processing based on the second-order statistic of received signal is employed. The use of higher-order statistic in signal processing is presented in this work, where the performance enhancement is evaluated in terms of an ultimate's array detection capability that represents the system ability to correctly determine the total number of sources present in an environment. The analysis uses the differential geometry as a tool to analyze the geometrical shape of the corresponding diversely polarized manifold for different orders of statistic. Its intrinsic parameter is then used for the derivation of the detection's lower bound. The theoretical framework is supported by computer simulations to examine the detection capability for various types of antenna arrays with different orders of statistic.
Read full abstract