Design and Realization of Band Pass Filter in K-Band Frequency for Short Range Radar Application

Short range radar (SRR) uses the K-band frequency range in its application. The radar requires high-resolution, so the applied frequency is 1 GHz wide. The filter is one of the devices used to ensure only a predetermined frequency is received by the radar system. This device must have a wide operating bandwidth to meet the specification of the radar. In this paper, a band pass filter (BPF) is proposed. It is designed and fabricated on RO4003C substrate using the substrate integrated waveguide (SIW) technique, results in a wide bandwidth at the K-band frequency that centered at 24 GHz. Besides the bandwidth analysis, the analysis of the insertion loss, the return loss, and the dimension are also reported. The simulated results of the bandpass filter are: VSWR of 1.0308, a return loss of -36.9344 dB, and an insertion loss of -0.6695 dB. The measurement results show that the design obtains a VSWR of 2.067, a return loss of -8.136 dB, and an insertion loss of -4.316 dB. While, it is obtained that the bandwidth is reduced by about 50% compared with the simulation. The result differences between simulation and measurement are mainly due to the imperfect fabrication process.

Bandwidth Analysis of Dual-Feed Slotted Antenna Using Artificial Neural Networks

In this chapter, artificial neural network is used for the estimation of bandwidth of a dual feed microstrip antenna. The MLPFFBP-ANN and RBF-ANN are used to implement the neural network model. The simulated values for training and testing the neural network are obtained by simulating the antenna on IE3D software. The results obtained by using ANNs and IE3D simulation are compared and are found quite acceptable, and also it is concluded that RBF network is more accurate and fast as compared to back propagation algorithm of MLPFFBP. The anticipated is applicable to operate in triple band from 2.208GHz-5.35GHz, 2.358GHz-2.736GHz, and 3.815GHz-5.143GHz. The antenna is also fabricated with FR-4 glass epoxy material. The experimental results, simulated results of IE3D, and simulated results of neural network are compared.