Figure 4 shows the S parameters of an unknown two-port device obtained with the use of the improved TSM technique and the SOLT method. This device consists of two coax-to-microstrip transitions, and two sections of 50-⍀ transmission lines with a series chip capacitor embedded in between. It is a suitable test device, because the magnitudes of its S parameters have a large variation over the measurement frequency range. From Figure 4 it can be seen that the proposed method is very accurate, especially for the phase measurements of S parameters.
DISCUSSION AND CONCLUSIONSThe conventional TSM and TOM methods were developed based on the linear calibration equations derived by Silvonen [7]. These two methods are very simple and do not have the phase uncertainty problem and any bandwidth limitation. For the reciprocal test fixtures, the conventional TSM method has four solutions. From Figure 2 one can see that at least one term obtained with any one of these four solutions is not accurate. The experimental results given in Figure 3 shows that the calibration accuracy can be significantly improved using the improved TSM method. It should be noted that for the improved TSM method, choosing any three equations among (1)-(4) will lead to the same solution, and the conventional and improved TSM methods are carried out with the same measurements. The only difference is using different algorithms. So, it is worth using the improved TSM method instead of the conventional TSM method.In principle, the calibration equations will become nonlinear when k is not regarded as an independent variable. When the short is considered as an imprecise standard, the improved TSM method gives a unique analytical solution. Therefore, the problem of phase uncertainty does not exist. From Figures 3 and 4 it can be seen that this improved TSM method has no bandwidth limitation.This analysis, mainly focused on the TSM method. Because the equation style obtained from the measurement with an open is the same as that with a short, the theory established can be applied to the TOM method. According to the present analyses, similar conclusions can be drawn. Kasten, Introducing the through-line deembedding procedure, IEEE MTT-S Int Microwave Symp Dig 1992, pp. 1455-1458. N.H. Zhu, Phase uncertainty in calibrating microwave test fixtures, IEEE Trans Microwave Theory Tech MTT-47 (1999MTT-47 ( ), 1917MTT-47 ( -1922 In the design of communication systems operating in a multipath environment, it is desirable to know the angular distribution of the incoming signal, in particular the angular spread. The angular spread can be directly related to fade duration, level-crossing rate, and coherence distance [1], and is a useful parameter in the design of diversity antenna systems for mobile radio. Through proper design of spatial and pattern diversity it is possible to reduce multipath fading and mitigate co-channel interference [2][3][4][5]. This Letter presents the direction-of-arrival (DOA) statistics in a typical urban environment extracted from both measuremen...