Yang Mingfei

Zero-bias model and compensation of VNA intermediate frequency system

A new zero-bias model and compensation procedure for the intermediate frequency (IF) signal processing system of the vector network analyzer (VNA) based on neural network is proposed in this paper. This procedure can be used to four IF channels with different conditions. The structure of the zero-bias compensation procedure in one port is given first. Then for four different channels, the same method is applied. The zero-bias is modeled by radial basis function neural network and the IF signal is compensated via the analysis based on the model. Finally, the proposed method is tested via AV3672 series VNA and the experiment shows that the method can well eliminate the zero-bias of the IF signal, and it is effective for zero-bias model and compensation.

Fault diagnosis of VNA intermediate frequency processing system based on dynamic fuzzy neural network

The paper presents a new fault diagnosis method for the intermediate frequency (IF) signal processing system of the vector network analyzer (VNA) based on dynamic fuzzy neural network (DFNN). This paper gives the structure of the fault diagnosis with three test points in one port first. Then for four different ports, it chooses the same method. The fault diagnosis is done by on-line self-organizing DFNN, and the structure and parameter identification is made in the on-line process without default values of structure and network parameters. It is the first time to introduce the DFNN into the fault diagnosis of the VNA IF system. Finally, the simulation experiment shows that the method can well approximate the nonlinear feature of the fault, and it is effective for fault diagnosis.