The two-antenna method is based on the Friis transmission equation and requires two identical samples of the antenna under test antenna: one acting as the radiating antenna, and the other as the receiving one.
The three-antenna method is used when only one sample of the test antenna is available. Then, any other two antennas can be used to perform three measurements, which allow the calculation of the individual gains of all three antennas. All three measurements are made at a fixed known distance between the radiating and the transmitting antennas.
The input impedance of an antenna is calculated via the reflection coefficient at its terminals, which are connected to a transmission line of known characteristic impedance. If the magnitude and the phase of the reflection coefficient are known, it is possible to calculate the antenna input impedance.
Because of real matching conditions, the gain of an antenna is reduced by the losses due to the mismatch of the antenna input impedance to the characteristic impedance. The gain obtained after this reduction is called realised gain.
In our case, the gain was measured using the VNA directly connected to the USLP 9143 log-periodic antenna under test, which was attached to a tripod on a lawn. The frequency range was calibrated between 700 and 2000 MHz in steps of 100 MHz with an intermediate-frequency bandwidth of 100 Hz and a stimulus output of +3 dBm.