Introduction
As the development of vehicle-to-everything (V2X) technology accelerates, antenna performance requirements for smart vehicles are becoming increasingly complex [
1,
2]. Antenna sensitivity is affected by the concentration of various electronic devices in small structures. Studies have explored how to improve antenna sensitivity through isolation from external factors because sensitivity directly influences the ability to receive signals [
3–
5]. However, electronic devices with internal antennas suffer greatly from onboard noise sources that are unintentionally generated by other electric components inside the printed circuit board (PCB). Moreover, onboard noise can be coupled with other resonances on the board, further reducing antenna performance.
This letter discusses the enhancement of antenna sensitivity through onboard noise suppression in an LTE/GPS communication module for V2X communication. The proposed module contains an equivalent noise circuit, which represents unexpected onboard noise, and two double-branch planar inverted-F antennas (PIFAs) targeting the B1 (2,100 MHz) and B5 (850 MHz) LTE communication bands and the L1 (1,575 MHz) and L5 (1,175 MHz) GPS communication bands. Each PIFA contains two radiation branches with different resonance frequencies sharing an identical feed loop, which exists for impedance matching [
6]. In this application, a noise circuit with a resonance frequency of 850 MHz can couple with the feed loop of the GPS antenna, and the magnitude of the coupling will be stronger as the resonance frequency of the feed loop approaches that of the operating band. For this reason, a noise suppression method must be implemented to avoid antenna sensitivity degradation due to the coupling of the noise sources.
Module Design and Noise Suppression
Fig. 1 illustrates the overall communication module and the simulated model with a specific antenna structure and an equivalent noise circuit. The module was set to be 100 mm × 50 mm with a 1 mm-thick FR4 substrate (
ɛr = 4.4), which is a commonly used vehicle tracker size. The simulation was conducted using the Ansys High-Frequency Structure Simulator (HFSS) tool in the driven modal solution type. The double-branch PIFA was designed in the upper left corner to operate in the B1 and B5 LTE communication bands, where
LL1 = 18 nH,
LL2 = 5 nH, and
CL = 2.2 pF. The branch containing
LL1 was set to adjust the resonance frequency to 850 MHz to operate in the B5 band. The magnitude of
CL can control the input impedance as the resonance frequency of the feed loop gets closer to the radiation branch, creating greater coupling.
An additional radiation branch containing
LL2 was inserted to operate in the B1 band. Another double-branch PIFA was designed in the upper right side to operate the L1 and L5 GPS communication bands, where
LG1 = 9 nH,
LG2 = 23 nH, and
CG = 1.0 pF. This was designed in mirror symmetry to the LTE antenna and with an identical operating mechanism. A noise equivalent circuit was inserted on the right side of the ground plane, where the capacitor
CN = 1.3 pF was used to set the resonance frequency of the B1 band. By defining the LTE antenna as port 1, GPS antenna as port 2, and noise circuit as port 3,
Z13′ can be written as follows [
7]:
where
Z13′ and Z13 are the mutual impedances with and without the GPS antenna, Z12 and Z23 are the mutual impedances between the GPS antenna and the LTE antenna and the noise circuit, respectively, and Z22 and ZL are the self-impedance and terminated load impedance at the GPS antenna. Z13 is a negligible value because the LTE antenna and the noise source are isolated. By comparing
Z13′ and Z13, the effect of the noise source coupling with the LTE antenna through the reference GPS antenna and the modified GPS can be estimated. If the resonance frequency of the feed loop of the GPS antenna approaches that of the noise circuit, Z22 + ZL approaches zero, and
Z13′ is strongly impacted. In other words, antenna sensitivity degradation is exacerbated by the coupling of the noise source and the GPS antenna. To suppress the sensitivity degradation effect, it is necessary to isolate the coupling between the GPS feed loop and the noise circuit by tuning the resonance frequency.
Fig. 2 shows the simulated current distribution of the GPS antenna with different capacitance values. When
CG = 1 pF, the resonance frequency of the feed loop approached 850 MHz, and the magnitude of the loop resonance current was strengthened. This resulted in the maximization of
Z12Z23/(
Z22 +
ZL), where
Z22 +
ZL approached zero. When
CG = 0.05 pF, the magnitude of
Z22 +
ZL significantly increased, making
Z12Z23/(
Z22 +
ZL) negligible value.
Fig. 3 represents the simulated reflection coefficient comparison of the LTE antenna for the cases of
Cg = 1.0 pF and
Cg = 0.05 pF in the GPS antenna and without an equivalent noise circuit. A resonance of the feed loop of the GPS antenna that could be coupled with the equivalent noise circuit antenna was observed when
Cg = 1.0 pF. The reflection coefficient was fairly close to the case without a noise circuit when
Cg = 0.05 pF. By tuning the resonance frequency of the feed loop, noise coupling could be avoided.
Fig. 4 shows the simulated reflection coefficient and the measured total efficiency comparison of the GPS antenna between the cases of
Cg = 1.0 pF and
Cg = 0.05 pF. Changing the capacitance value led to an impedance mismatch, which resulted in a degradation of the total efficiency. However, the difference in the peak total efficiency between the two cases was less than 8%, which is not crucial for GPS antenna performance.
Table 1 shows the active test results, including the measured total radiated power (TRP) and total isotropic sensitivity (TIS) values in dBm for the proposed LTE antenna. TRP is the sum of all power radiated by an antenna, and TIS refers to the lowest detectable received signal strength. These values were calculated as follows [
8]:
where generally used units of TRP, TIS, Pcond, Scond, Sinterference are in dBm, and efficiency is in dB. Pcond and Scond, the conducted power and sensitivity from the commercial IC chip, were 22 dBm and −99 dBm, respectively, for the proposed application. These results imply that the IC chip generates a signal level of 22 dBm to the antenna input in transmitting mode and recognizes the signal, which is stronger than −99 dBm, in receiving mode. It is noted that lower TIS leads to better antenna sensitivity. Sinterference is determined by noise coupling within the onboard noise source. For the TIS data comparison at the B5 band, the coupled noise source was suppressed by tuning the resonance frequency of the feed loop of the GPS antenna. In other words, Sinterference was reduced, resulting in a TIS improvement of more than 12 dB.