No abstract

No abstract

simple example illustrates the use of our equations in an airborne radar application. We can explain the results in Figs. 1-7 by considering the changes in the pdf of caused by changes in Var fdg or C: In these cases, a decrease in Var fdg or j causes the mass in the pdf of to move toward larger values of : Due to the decreasing nature of the function multiplying the pdf of in the integrand of (30), this causes a decrease in the probability of false alarm. It is possible to apply similar analysis to also explain changes in probability of detection. VI. CONCLUSIONS An analysis of the performance of the adaptive matched filter algorithm has been provided for cases where the data used to estimate the covariance matrix is not matched to the true covariance matrix of the data to be tested. Such cases can occur in nonhomogeneous environments that appear to occur frequently in real radars. Closed-form approximate expressions are given for the probability of false alarm and detection. These expressions apply for any amount of data used in the covariance matrix estimation. The analysis indicates which types of covariance matrix mismatches are important and which types are not. The equations indicate that performance depends on a few critical parameters. An airborne radar example is provided to show that the changes in performance due to mismatch can be significant in some practical situations. REFERENCES [1] W. S. Chen and I. S. Reed, "A new CFAR detection test for radar," Abstract-Multiresolution ESPRIT is an extension of the ESPRIT direction finding algorithm to antenna arrays with multiple baselines. A short (half wavelength) baseline is necessary to avoid aliasing, and a long baseline is preferred for accuracy. The MR-ESPRIT algorithm allows the combination of both estimates. The ratio of the longest baseline to the shortest one is a measure of the gain in accuracy. Because of various factors, including noise, signal bandwidth, and measurement error, the achievable gain in accuracy is bounded.

Reduction of propagation loss of terahertz graphene plasmon can be made by increasing the chemical potential of graphene layer, but at the cost of significantly increased modal area, which fundamentally limits the packing density on a chip. By utilizing the strong coupling between the dielectric waveguide and plasmonic modes, we propose hybrid plasmonic terahertz waveguides that not only significantly suppress the mode field confinement, but also maintain a compact modal size. A typical propagation length is 127 μm, and optical field is confined into an ultra-small area of approximately 32.6 μm 2 at 3 THz. This structure also exhibits ultra-low crosstalk, which shows great promise for constructing various functional devices in future terahertz integrated circuits.