1.

INTRODUCTION

The cable network plays a critical role in communicating between sensors of control system in metallic environment such as vehicle, spacecraft and industrial equipment^1-3. With the increasing application of various sensors, the complexity of the harness connection is concerned^4-7. However, in view of the ubiquitous requirements of non-line-of-sight (NLOS) transmission with strong fading caused by various metal components in narrow, compact and closed metal spaces, traditional radio frequency (RF) or optical wireless communication technology is faced with great challenges^8-10. As a solution to this problem, the method of capacitive coupling communication within a metal cabinet is becoming an appealing technology^{7, 11, 12}. Capacitive coupling communication was firstly proposed by Zimmerman¹³, and was widely used in human body communication to realize the data transmission between wearable devices or implantable devices using electric field coupling^14-16.

As far as we know, few works conducted in a metal cabinet to realize NLOS capacitive coupling communication, and the modeling and characterization have not been systematically investigated. Thus in this paper, focusing on the frequency range of 1-100 MHz, we attempt to derive the propagation characteristics under different electrode parameters of the capacitive coupling communication channel inside a metal cabinet. The composition and structure of the communication system are proposed. A finite element simulation model is built in ANSYS Maxwell software and the effect of electrode parameters such as electrode height, area and distance on transmission gain is simulated. Furthermore, experimental measurements are carried out based on a vector network analyzer to further verify the correctness of the simulation results.

2.

COMMUNICATION SYSTEM STRUCTURE

Unlike body area communication, which takes the human body as the forward channel and the ground as the backward channel, in a closed metal cabinet such as aerospace vehicles, there is only one circuit of the shell without a direct ground circuit, and the metal body may be coupled with the electric field in the direction of 360 degrees. Therefore, we spray two conductive layers on the surface of the metal body, one as the forward channel and the other as the backward channel, so as to form a complete signal transmission circuit. The schematic diagram of the capacitive coupling communication system in a closed metal cabinet is shown in Figure 1. The figure is described from the top view inside the cabinet. The orange part is the inner surface of the metal cabinet as the application environment of the communication system. The two parallel green cuboids adjacent to each other are the forward and backward metal channels laid on the surface of the metal body to form a differential pair for transmitting electric field signals. The red squares in the figure are the transceiver electrodes of the communication nodes, which are electrically insulated from the channel surface and spaced at a certain height for the coupling induction of quasi-electrostatic field signals. The transmitter applies low-frequency voltage signal to the transmitting electrode to generate a quasi-electrostatic field near its surface. When the charge of transmitting plate 1 is positive and the charge of plate 2 is negative, opposite charges are induced on the forward and backward channel surfaces, respectively. In order to reach the static equilibrium state of the channel as an isolated conductor, the Coulomb force makes the charge away from or close to the electric field source, causing the charge flow on the surface of the channel, forming an induced charge flow circuit between conductors 1-7 to realize the propagation of quasi-electrostatic field signal.

Figure 1.

The schematic diagram of the communication system.

Each pair of electrodes is composed of two electrodes with opposite polarity, which can be simplified into an electric dipole model for research. The channel is a plane conductor plate without electrical connection, which can be equivalent to an isolated conductor model. Therefore, the electric field generated by the transmitting electrodes in the cabinet can be studied by using the model of horizontally placing the electric dipole above the plane conductor plate and using the mirror image method. The potential induced by the electric dipole on the channel surface of the transmitting end is as follows:

Then the normal electric field intensity on the channel surface of the transmitting end is:

where p is the electric dipole moment; h is the distance between the electrode and the channel surface; x, y and z are the position coordinates of the studied point, respectively. The charge density induced on the channel surface of the transmitting end is as follows:

Then the total reverse charge on the channel surface of the transmitting end is:

where w_s is the width of the metal channel; l_s is the length of the metal channel in the transmission section; y₀ is the central coordinate of the metal channel in the y direction. Thus, according to equations (1) and (4), the normal electric field intensity on the channel surface of the receiving end is obtained:

where h_r is the distance between the channel and the cabinet wall; x_r, y_r and z_r are the position coordinates of the studied point, respectively; w_r is the width of the metal channel; d_r is the spacing of metal channels. It can be found from the formula that the transmission process of induced electric field from the transmitting end to the receiving end is affected by electrode parameters and channel parameters. In order to reduce the transmission loss, these parameters need to be optimized.

3.

SIMULATION MODEL

FEM simulation model is shown in Figure 2. It mainly includes the metal cabinet, isolator, transmission channels and transceiver electrodes. The cabinet is a metal cavity with a length of 3 m, a diameter of 1.2 m and a thickness of 1 mm, and is divided into two cavities by an isolator with a length of 1m, which is used to simulate the non-line-of-sight communication between transceiver nodes. The cabinet material is set as aluminium, the relative dielectric constant is set to 1, and the conductivity is set to 5.8E+7 S/m. The transmission channels are two long metal strips with a length of 3 m, a width of 0.1 m and a thickness of 1 mm, which are laid parallel on the surface of the cabinet wall to transmit electric field signals. The channel material is aluminium, the relative dielectric constant is set to 1, and the conductivity is set to 5.8E+7 S/m. In order to ensure the electrical insulation between the channel and other conductors, insulating layers with thickness of 1 mm are laid on the upper and lower surfaces of the metal channels, and the dielectric constant is the same as that of the conductive layers. The transmitting and receiving electrodes are separated in two cavities. The positive electrodes are parallel and suspended in the forward channel, and the negative electrodes are parallel and suspended in the backward channel. The size of all electrodes is set to 0.1 m × 0.1 m × 0.01 m, and the material is set as copper.

Figure 2.

FEM simulation model.

Figure 3 shows the electric field simulation distribution in the metal cabinet with or without metal channels. It is confirmed that when there is no metal channel, the electric field generated at the transmitting end cannot be transmitted to the receiving end through the isolator. When two metal channels are arranged, a complete signal transmission loop is formed, and the electric field signal at the transmitting end can be well transmitted to the receiving end. In addition, when the channel length is shortened, the electric field is only distributed in the section with channel laying. Therefore, the electric field is only distributed near the channel surface and cannot be transmitted to the part without channel laying.

Figure 3.

Electric field distribution.

4.

COMPARISON OF TRANSMISSION CHARACTERISTICS

4.1

Simulation effect of electrode distance

Under the condition of fixed transceiver electrode area of 0.01 m² and transceiver electrode height of 1 cm, the path loss characteristics under different transceiver electrode distance are simulated and analyzed. The transceiver electrode distances of 2.7 m, 2.5 m, 2.3 m, 2.1 m, 1.9 m and 1.7 m are selected. The simulation results are shown in Figure 4.

Figure 4.

Simulation channel gain under different electrode distance.

It can be seen that under the condition of a certain electrode area and electrode height, the channel gain is not affected by the transceiver electrode distance. That is because changing the transceiver electrode distance will not change the coupling capacitance between the electrode and the channel, the capacitance parameters in the equivalent circuit model and the channel transfer function remain unchanged. Therefore, the path loss is independent of the distance between the transmitting and receiving electrodes. In addition, the channel gain increases with the increase of signal frequency, which reflects the high pass characteristics of signal transmission in the certain frequency range.

4.2

Simulation effect of electrode area

Under the condition of fixed transceiver electrode distance of 2.3 m and transceiver electrode height of 1 cm, the path loss characteristics under different transceiver electrode area are simulated and analyzed. The transceiver electrode areas of 0.0144 m², 0.01 m², 0.0064 m², 0.0036 m², 0.0016 m² and 0.0004 m² are selected. The simulation results are shown in Figure 5.

Figure 5.

Simulation channel gain under different electrode area.

It can be seen that under the condition of a certain transceiver electrode distance and transceiver electrode height, the channel gain decreases with the decrease of transceiver electrode area. In particular, when the electrode area is reduced to 0.0004 m², the attenuation of transmission signal increases greatly. This is because reducing the electrode area reduces the coupling capacitance between the electrode and the channel, which changes the capacitance parameters and channel transfer function in the equivalent circuit model. Therefore, the path loss is related to the transceiver electrode area.

4.3

Simulation effect of electrode height

Under the condition of fixed transceiver electrode distance of 2.3 m and transceiver electrode area of 0.01 m², the path loss characteristics under different transceiver electrode height are simulated and analyzed. The transceiver electrode heights of 1 cm, 3 cm, 5 cm, 7 cm and 9 cm are selected. The simulation results are shown in Figure 6.

Figure 6.

Simulation channel gain under different electrode height.

It can be seen that under the condition of a certain transceiver electrode distance and transceiver electrode area, the channel gain decreases with the increase of transceiver electrode height. In particular, when the electrode height increases from 1 cm to 3 cm, the transmission signal attenuates greatly. This is because increasing the electrode height reduces the coupling capacitance between the electrode and the channel, which changes the capacitance parameters and channel transfer function in the equivalent circuit model. Therefore, the path loss is related to the height of the transceiver electrode.

5.

EXPERIMENTAL VALIDATION

5.1

Experimental setup

The test platform of frequency-domain characteristics is shown in Figure 7. The test is carried out through a portable vector network analyzer (NanoLab SAA-2N) to measure the path loss curve (S21 parameter curve) of the capacitive coupling communication channel. Before testing, the instrument must be strictly calibrated using the corresponding calibration elements to ensure the measurement accuracy, including short-circuit calibration, open-circuit calibration and load calibration between the two ports. The port 1 and port 2 of the vector network analyzer are respectively connected with the transmitting electrode and the receiving electrode. The scanning is carried out point by point within the measurement bandwidth of 1~100 MHz. The number of frequency scanning points is 101. The transmission signal with determined power is transmitted to the transmitting electrode through a low loss cable for transmission. The signal is coupled to the metal channel and received by the receiving electrode through the transmission channel. Finally, the signal is transmitted back to the vector network analyzer through another low loss cable, and the signal frequency-domain attenuation characteristic from transmitter to receiver is obtained. The vector network analyzer is connected with the host computer through USB interface, and the measured curve is saved using the corresponding software in the host computer.

Figure 7.

Experimental setup.

5.2

Experimental effect of electrode distance

In order to study the effect of transceiver electrode distance on the channel transmission attenuation, under the condition of using 4 cm × 4 cm copper electrodes and 1 cm electrode height, the signal transmission loss curves are measured when the distance between transmitting and receiving electrodes is 1 m, 0.8 m, 0.6 m and 0.4 m, respectively. The test results are shown in Figure 8.

Figure 8.

Experimental channel gain under different electrode distance.

The results show that the signal transmission attenuation basically unchanged with the change of transceiver electrode distance, which is consistent with the above simulation results. This is because changing the transceiver electrode distance does not affect the coupling capacitance between the electrode and the metal channel, so the transmission loss is basically not affected.

5.3

Experimental effect of electrode height

In order to study the effect of electrode height on the channel transmission attenuation, under the condition of using 4 cm × 4 cm copper electrodes and 1 m transceiver electrodes distance, the signal transmission loss curves when the electrode height is 1 cm, 2 cm, 3 cm and 4 cm are measured, respectively. The test results are shown in Figure 9.

Figure 9.

Experimental channel gain under different electrode height.

The results show that when the electrode height increases, the signal transmission attenuation increases accordingly, which is consistent with the above simulation results. This is because the coupling capacitance between the electrode and the channel decreases with the increase of the electrode height, so that part of the electric field diffuses into the environment and increases the signal transmission loss.

5.4

Experimental effect of electrode area

In order to study the influence of electrode area on the channel transmission attenuation, under the condition of using 1 m transceiver electrode distance and 1 cm electrode height, the signal transmission loss curves when the electrode area is 4 cm × 4 cm、3 cm × 3 cm、2 cm × 2 cm are measured, respectively. The test results are shown in Figure 10.

Figure 10.

Experimental channel gain under different electrode area.

The results show that when the electrode area increases, the signal transmission attenuation decreases accordingly, which is consistent with the simulation results. This is because increasing the electrode area increases the coupling capacitance between the electrode and the channel, so the electric field coupled to the metal channel increases, and the transmission loss decreases.

6.

CONCLUSION

In this paper, the channel transmission characteristics of capacitive coupling communication within a metal cabinet was investigated. The composition and structure of the communication system were proposed firstly, and the communication system was explained and analyzed theoretically. FEM simulation and experimental measurement were conducted under various electrode height, area and distance to study the effect of different electrode parameters on the channel transmission characteristics.

The finite element simulation results showed that the electric field at the transmitting end can be transmitted across the isolator to the receiving end with two metal channels laid. It was also found that the transmission gain of the system can be improved by reducing the electrode height and increasing the electrode area, while the distance between the transmitting and receiving electrodes has little effect on the path loss. A good agreement was showed between the simulation results and experimental results. The channel modeling and characterization of the capacitive coupling communication within a metal cabinet achieved in this paper can promote the design and optimization of the communication system in the future research.