Low-loss filtering transmission type reconfigurable intelligent surface

Abstract

The invention discloses a low-loss filtering transmission-type reconfigurable intelligent surface, and aims to solve the problems that an existing filtering reconfigurable intelligent surface is insufficient in frequency selectivity, high in insertion loss, narrow in beam coverage range and difficult in collaborative design of transmission and regulation. The structure adopts a six-layer metal laminated framework, a receiving layer adopts a symmetrical H-shaped patch and is embedded with a PIN diode to realize 1-bit phase modulation, a regulation and control layer and a transmitting layer cooperatively introduce double transmission zeros through an E-shaped open circuit branch knot, a symmetrical slot and a parasitic branch knot, and low-loss and high-selectivity filtering is realized by combining multi-physical field optimization. The working frequency band is 6.23-7.38 GHz, the 3dB relative bandwidth reaches 17%, the minimum transmission loss in a passband is-0.25 dB, the phase regulation error is smaller than + / -5 degrees, + / -70-degree wide-angle-domain dual-beam scanning is supported, and the out-of-band suppression depth breaks through-40 dB. The method integrates the functions of filtering, low loss, wide-angle beam forming and the like, is suitable for scenes of dynamic spectrum management, anti-interference transmission and the like in a 6G intelligent communication system, and has a wide application prospect.

Description

Technical Field

[0001] This invention relates to the fields of mobile communication and electromagnetic control technology, specifically to a low-loss filtered transmission type reconfigurable smart surface. It is particularly suitable for dynamic spectrum shaping, anti-interference communication, and wide-angle beamforming scenarios in sixth-generation (6G) mobile communication systems. It can be applied to cognitive radio, dense heterogeneous networking, and integrated sensing and communication systems, where high spectrum utilization and communication reliability are required. Background Technology

[0002] As 6G mobile communication systems evolve towards intelligent wireless environments and dense heterogeneous networks, the requirements for spectral efficiency, communication reliability, and scenario adaptability in wireless communication continue to increase. Reconfigurable smart surfaces (RIS), as a key enabling technology for dynamically controlling electromagnetic wave propagation characteristics, can achieve beamforming, dynamic spectrum management, and physical layer security enhancement through programmable amplitude and phase modulation, becoming a core solution to overcome traditional channel bottlenecks. However, traditional RIS research has largely focused on spatial domain characteristic control, such as amplitude and phase modulation mechanism optimization and beam scanning range expansion, lacking sufficient ability to control frequency domain characteristics. It lacks functions such as bandpass response optimization, stopband suppression, and frequency selective control, making it difficult to meet the anti-interference communication and high spectrum utilization requirements of complex 6G scenarios. This shortcoming is gradually becoming a key bottleneck restricting the development of RIS in 6G.

[0003] To achieve precise control of the electromagnetic spectrum, filtering RIS (Responsive Frequency Selective Surface) technology has emerged. It improves signal purity and suppresses out-of-band interference through frequency selective response control, making it irreplaceable in scenarios such as 6G low-altitude UAV communication and dense heterogeneous networking. In the technological evolution of filtering RIS, early research focused on frequency selective surfaces (FSS), using metal patches or aperture units on a dielectric substrate to achieve filtering functions in specific frequency bands. These were widely used in stealth radar radomes, satellite communication subreflectors, and other fields. However, traditional FSSs are passive devices with fixed resonant frequencies and phase responses, lacking adaptability to the external electromagnetic environment and unable to meet the dynamic requirements of 6G scenarios. To address this issue, reconfigurable frequency selective surfaces (RFSS) were proposed. By introducing active devices such as PIN diodes and varactor diodes or tunable materials, resonant frequency tuning and passband switching were achieved. However, RFSS still cannot achieve beamforming and lacks spatial domain signal processing capabilities, making it difficult to meet the high-precision beam manipulation requirements of 6G.

[0004] With technological advancements, filtering RIS, as an advanced form of FSS and RFSS, has gradually acquired integrated capabilities for spectrum modulation and beamforming. However, existing solutions still suffer from numerous technical shortcomings. In the field of transmissive filtering RIS, while some designs exhibit a degree of frequency selectivity, the filter chip and multi-cascaded structure result in high insertion loss (e.g., the current technology has an in-passband insertion loss of 2.5dB), and strong inter-unit coupling limits the beam scanning range (typically not exceeding 30°). While reflective filtering RIS demonstrates some frequency selectivity, its phase modulation accuracy is low (only 180°±20°), bandwidth is narrow (1dB operating bandwidth is only 7.5%), and out-of-band rejection capability is limited (typically greater than 10dB), failing to meet the requirements of strong interference scenarios. More importantly, transmissive RIS itself faces stringent technical challenges: it must simultaneously achieve low signal transmission loss through the medium and precise phase modulation across interfaces, involving complex electromagnetic field transmission-modulation collaborative design. Existing solutions struggle to balance transmission efficiency, phase response, and filter roll-off characteristics, limiting their application in physically isolated communication and transparent environment adaptation scenarios.

[0005] In summary, current filtering RIS systems generally suffer from insufficient frequency selectivity, high insertion loss, narrow beam coverage, and low phase modulation accuracy. In particular, the technical bottleneck of transmission-type filtering RIS systems in the coordinated optimization of multiple performance indicators severely restricts their application in 6G intelligent communication systems. Therefore, developing a transmission-type RIS system that combines low loss, high selective filtering, broadband response, and wide-angle beam scanning capability has become an urgent need to solve the challenges of spectrum management and efficient communication in complex 6G scenarios. Summary of the Invention

[0006] Technical issues:

[0007] This invention aims to solve the problems of insufficient frequency selectivity, high insertion loss, narrow beam coverage, and difficulty in coordinating transmission and modulation of existing filtered RIS.

[0008] Technical solution:

[0009] To address the aforementioned technical issues, this invention proposes a low-loss filtered transmission reconfigurable smart surface (LF-TRIS), which employs a six-layer metal stack architecture and integrates structural design, device selection, and model optimization to achieve synergistic improvement in multiple performance aspects.

[0010] The core design includes: the receiving layer uses a symmetrical H-shaped patch with symmetrically embedded MACOM MADP-000907-14020x PIN diodes, and 1-bit phase modulation is achieved by switching between two bias states; the control layer is equipped with E-shaped open-circuit stubs, and the transmitting layer is equipped with symmetrical slots and parasitic stubs, which work together to introduce dual transmission zeros; a hybrid coupling model of controllable dual transmission zeros and three reflection poles is constructed, and combined with multi-physics field collaborative optimization, to achieve a balance between low loss and high selective filtering.

[0011] The DC feed layer adopts a fan-shaped patch structure to isolate the radio frequency signal from the DC bias and avoid crosstalk; both the receiving layer and the transmitting layer adopt F4B dielectric substrate, and the control layer is sandwiched between the first reference ground layer and the second reference ground layer. Multi-layer electromagnetic coupling is achieved through metallized vias to reduce ohmic loss.

[0012] This scheme achieves a 17% relative bandwidth of 3 dB in the 6.23–7.38 GHz band, a frequency selectivity coefficient of 0.72, an out-of-band rejection depth exceeding -40 dB, a minimum transmission loss of -0.25 dB in the passband, a phase modulation error of less than ±5°, and supports dual-beam scanning in a wide-angle domain of ±70°. Wireless communication experiments have verified that it can achieve high-quality signal transmission at the passband frequency and effectively suppress interference at the stopband frequency, meeting the requirements of practical communication scenarios.

[0013] Preferably, the feeder layer is designed with a 60° fan-shaped structure to optimize impedance matching and reduce signal reflection;

[0014] Preferably, in the low-loss design, the receiving and transmitting patch layers of the metasurface use F4B low-loss dielectric (tanδ=0.0015) between them and the adjacent reference ground layer; Rogers RO4450F (tanδ=0.004) is used between the first reference ground layer and the E-shaped open stub, and between the second reference ground layer and the DC feed layer; Rogers RO4350B (tanδ=0.0037) is used between the second reference ground layer and the DC feed layer; the cell size is 13.17mm × 11.64mm (approximately 0.29λ × 0.25λ, where λ is the wavelength of 6.6GHz).

[0015] The specific steps include:

[0016] Step 1: Cell Design: The LF-TRIS cell is designed using a six-layer metal stack-up architecture, including a receiver layer, a first reference ground layer, a control layer, a second reference ground layer, a DC feed layer, and a transmitter layer. The receiver and transmitter layers use F4B dielectric substrates (dielectric constant 2.65, loss tangent 0.0015), with symmetrical H-shaped and U-shaped patches etched respectively. The control layer is sandwiched between the first reference ground layer (using Rogers RO4450F material) and the second reference ground layer (using Rogers RO4350B material), with an E-shaped open-circuit stub. The DC feed layer uses a fan-shaped patch to achieve RF and DC isolation. Symmetrical slots are cut into the H-shaped patch of the receiver layer, embedding two MACOM MADP-000907-14020x PIN diodes. Switching between the ±2.5V bias enables two operating states, ensuring mirror symmetry of the surface current and achieving 180°±5° 1-bit phase modulation. See attached diagram for details. Figure 4 As shown, the key dimensional parameters are listed in the table below:

[0017]

[0018] Step 2: Filter Optimization: Dual transmission zeros are introduced through structural co-design. Symmetrical slots and parasitic stubs are incorporated into the U-shaped emitter patch, while E-shaped open-circuit stubs are arranged in the control layer to control the zero positions at low frequencies (5.93 GHz) and high frequencies (7.69 GHz), respectively. A low-loss optimization strategy is employed, using metallized vias to construct magnetic coupling paths and selecting low-series-resistance (0.25 Ω) PIN diodes to enhance out-of-band energy blocking. A controllable dual transmission zero and three-reflection pole hybrid coupling model is established. Through multi-physics co-optimization, a 3 dB relative bandwidth of 17%, a frequency selectivity coefficient of 0.72, and an out-of-band rejection depth exceeding -40 dB are achieved.

[0019] Step 3: Array Integration and Control: An array is constructed using 8×8 cells as the basic module. Multi-layer electromagnetic coupling is achieved through metallized vias, and four modules are spliced ​​together to form a large 16×16 array. An FPGA control board is used to build the control system, connecting the array feed layer via FPC flexible cables. Upper computer command transmission is implemented based on the UDP protocol. 1-bit digital encoding is converted into voltage signals, and each PIN diode is independently biased and controlled, achieving fast switching between state 0 (+2.5V) and state 1 (-2.5V), solving the problem of dense feeding in large-scale arrays and supporting real-time beamforming.

[0020] Step 4: S-parameter testing: A darkroom lens testing system was set up, connecting the transmitting and receiving horn antennas to a Keysight E5071C vector network analyzer, covering the test frequency band of 5.5-8.0 GHz. Absorbing material was filled around the array to prevent beam leakage, and the S-parameters were tested for two operating states. In the measured passband of 6.23-7.38 GHz, the lowest transmission loss reached -0.25 dB, the reflection coefficient was below -10 dB, and the phase error was ≤ ±5°. The low-frequency null suppression depth was -40.67 dB at 5.72 GHz and -23.62 dB at 7.6 GHz. The null frequency offset was calibrated using methods such as dielectric parameter compensation.

[0021] Step 5: Dual-beam far-field pattern test: The test was conducted in a microwave anechoic chamber. The transmitting antenna was 1.5 meters away from the sample, and the receiving antenna was 5 meters away and connected to a spectrum analyzer. The sample and transmitting antenna rotated with the turntable. Using 6.8 GHz as the center frequency, a column control code was applied to the 8×8 array, and the transmitted signal at different angles was tested. The measured results show that the dual-beam scanning angle covers ±70°, and stable beam focusing is achieved at angles of ±18°, ±35°, ±55°, and ±70°, verifying the wide-angle domain control capability.

[0022] Step 6: Communication Experiment Verification: Two USRP-2974 software-defined radios were used as transceivers. The transmitter encoded the color video sequence into a binary data stream, which was then QPSK modulated and loaded onto the RF carrier. A 5-8GHz broadband horn antenna was selected as the transceiver antenna, with an absorbing screen with a window placed in the center. An LF-TRIS was embedded in the center of the window, and the transceiver antenna was 0.8 meters from its normal direction. Signal transmission at 5.72GHz (stopband), 6.8GHz (passband), and 7.6GHz (stopband) was tested. The video was clearly demodulated and the constellation diagram converged within the passband, while the signal was severely distorted within the stopband, verifying the filtering and anti-interference performance.

[0023] Beneficial effects:

[0024] 1. By adopting a multi-physics collaborative optimization design using a low-loss dielectric substrate (F4B), magnetic coupling path (metallized via), and low series resistance PIN diode, ultra-low passband insertion loss is achieved, as low as -0.25 dB, which significantly improves energy transmission efficiency and reduces system energy consumption.

[0025] 2. Through the collaborative structural design of the receiving layer, the control layer and the transmitting layer, especially by using the coupling of E-shaped open-circuit stubs and symmetrical slots to generate high-frequency transmission zeros and the independent generation of low-frequency transmission zeros by parasitic stubs, a controllable dual transmission zero filtering mechanism is constructed, which enables this surface to have excellent high frequency selectivity (coefficient up to 0.72) and strong out-of-band suppression capability (depth exceeding -40 dB), effectively ensuring the purity of the signal within the passband.

[0026] 3. A large-scale array based on 1-bit phase modulation unit (H-shaped patch and symmetrical PIN diode), combined with the independent addressable bias technology of the underlying FPGA control board, achieves high-precision (phase error < ±5°) and wide-angle domain (covering ±70°) dual-beam scanning and shaping capability, overcoming the shortcomings of the narrow beam coverage of existing transmission-type RIS.

[0027] 4. The innovative DC-DC feed layer (fan-shaped patch) design effectively isolates the RF signal from the DC bias path, avoiding crosstalk and ensuring stable low-reflection characteristics within the operating frequency band (S). 11 <-10 dB).

[0028] 5. This invention integrates high-selectivity filtering, ultra-low loss transmission, broadband response (17% relative bandwidth), and wide-angle beamforming into a single transmission-type RIS hardware platform, solving the technical challenge of synergistic optimization of transmission and modulation performance, and providing an efficient hardware solution for applications such as dynamic spectrum shaping and anti-interference transmission in 6G smart communication. Attached Figure Description

[0029] Figure 1 This is a conceptual diagram of the LF-TRIS filter in this invention. The out-of-band zeros and in-band frequencies are selected and tested in an anechoic chamber to measure their far-field radiation patterns. 21 In the parameter diagram;

[0030] Figure 2 In this invention, the current distribution on the surface of the top-layer patch is shown in two states: state 0: PIN1 is reverse biased and PIN2 is forward biased; state 1: PIN1 is forward biased and PIN2 is reverse biased.

[0031] Figure 3 Simulation results of the normal incident S-parameters of the designed LF-TRIS under two states: (a) S 21 and S 11 (a) Amplitude; (b) Phase and phase difference;

[0032] Figure 4 The unit structure of the designed LF-TRIS: (a) H-shaped patch of the receiving layer of the LF-TRIS, with slotted embedded diodes PIN1 and PIN2; (b) E-shaped open stub of the control layer; (c) DC feed layer, used to isolate RF signals and connect DC voltage; (d) exploded view of the LF-TRIS; (e) U-shaped patch of the transmitting layer;

[0033] Figure 5The zero-point design process for the designed LF-TRIS: (a) Cell1: First stage, only symmetrical slot structure is placed; (b) Cell2: Second stage, symmetrical slot and E-shaped open-circuit stub are placed; (c) Cell3: Third stage, symmetrical slot, E-shaped open-circuit stub and parasitic stub are placed; (d) Surface current of symmetrical slot in Cell1; (e) Surface current of E-shaped open-circuit stub in Cell2; (f) Surface current of parasitic stub in Cell3; (g) S-parameter amplitude simulation of Cell1; (h) S-parameter amplitude simulation of Cell2; (i) S-parameter amplitude simulation of Cell3;

[0034] Figure 6 In this invention, the S of the three types of units 21 Parametric simulation results: Initial Cell, transmission unit without filter structure; Unoptimized Filtering Cell, transmission unit with filter structure but without optimization; Optimized Filtering Cell, transmission unit with filter structure and parameter co-optimization.

[0035] Figure 7 In this invention, the three control methods for normal incident S 21 Parameter simulation results: (a) Adjusting L7; (b) Adjusting L3; (c) Adjusting L9;

[0036] Figure 8 The LF-TRIS experimental prototype and its functions in this invention are as follows: an 8×8 reconfigurable cell array (64 cells in total), multi-layer electromagnetic coupling is achieved through metallized vias; 128 PIN diodes are integrated, and 1-bit phase modulation is achieved through symmetrical bias switching; flexible printed circuit (FPC) cables are connected to the underlying FPGA control board to support real-time encoding and reconfiguration;

[0037] Figure 9 The following is a schematic diagram of the lens test for the LF-TRIS prototype in this invention and its test scenario: (a) Schematic diagram of the lens test for the LF-TRIS prototype: The transmitting end outputs an RF signal from a vector network analyzer to a standard gain horn, which is then incident vertically onto the center of the RIS array. The receiving end receives the signal from the horn antenna at the focal point and transmits it back to the vector network analyzer to analyze the S-parameters. The FPGA transmits state 0 or state 1 encoded to the RIS array PIN diode; (b) Actual lens test scenario: The prototype is a 16×16 reconfigurable unit (assembled from four 8×8 daughterboards), and the test frequency band covers 5.5–8.0 GHz;

[0038] Figure 10The S-parameter characteristic curve of LF-TRIS in this invention is as follows: (a) S 21 and S 11 Amplitude analysis: The effectiveness of the full-wave simulation model was verified, and the wide bandwidth, low loss, high frequency selectivity and strong out-of-band suppression of LFTRIS were fully demonstrated; (b) Phase difference between the two states: The phase difference in the 6.27-7.42 GHz operating frequency band was less than ±5° compared with 180°.

[0039] Figure 11 The following are schematic diagrams and actual test results of the LF-TRIS sample in this invention: (a) Schematic diagram of far-field radiation pattern test; (b) Actual test result of far-field radiation pattern;

[0040] Figure 12 The following are the measured results of far-field dual-beam scanning of the LF-TRIS sample in this invention: (a) θ = ±18°; (b) θ = ±35°; (c) θ = ±55°; (d) θ = ±70°;

[0041] Figure 13 This is a schematic diagram illustrating the configuration of a wireless communication experiment based on an LF-TRIS prototype in this invention. An absorbing window is placed between the transmitting and receiving antennas, and an LF-TRIS is embedded within the window.

[0042] Figure 14 The QPSK wireless communication experiments in this invention are as follows: (a) a measured display of a frame of wireless video transmission of a 5.72 GHz signal; (b) a measured constellation diagram of wireless video transmission of a 5.72 GHz signal; (c) a measured display of a frame of wireless video transmission of a 6.8 GHz signal; (d) a measured constellation diagram of wireless video transmission of a 6.8 GHz signal; (e) a measured display of a frame of wireless video transmission of a 6.8 GHz signal; and (f) a measured constellation diagram of wireless video transmission of a 7.6 GHz signal. Detailed Implementation

[0043] This invention provides a low-loss 1-bit transmissive reconfigurable smart surface (LF-TRIS) that integrates highly selective filtering and wide-angle beamforming. The core of this invention lies in its multi-layer coupling structure and dynamic control mechanism, which integrates electromagnetic wave spectrum shaping, low-loss transmission, and wide-angle beam scanning. The specific implementation of this invention is described in detail below with reference to the accompanying drawings, progressing from the filtering control principle, unit structure design, simulation optimization, to experimental verification.

[0044] First, refer to Figure 1The LF-TRIS filter concept diagram shown illustrates that this invention requires the construction of a frequency selection and beam focusing synergy model. Far-field tests were conducted using a passband center frequency of 6.8 GHz, an out-of-band low-frequency zero of 5.72 GHz, and a high-frequency zero of 7.6 GHz. Under 0° coding conditions, the gain in the 0° direction within the passband reached 18.38 dB, while the gains at the out-of-band zeros decreased to 10.67 dB and 8.15 dB, respectively, verifying the synergistic characteristics of frequency selection and beam focusing. The core of this model is to achieve low-loss transmission within the passband and strong out-of-band suppression through the hybrid coupling effect of dual transmission zeros and three reflection poles.

[0045] Next, a 1-bit transmission unit structure was designed to achieve phase modulation and filtering functions. For example... Figure 2 As shown, the unit switches between two operating states via a dual-PIN diode (MACOM, MADP-000907-14020x): state 0 (PIN1 reverse biased, PIN2 forward biased) and state 1 (PIN1 forward biased, PIN2 reverse biased). The surface current of the top layer is mirror-symmetrical in both states, ensuring a 180° phase difference. Figure 3 Simulation results show that the phase error is ≤ ±5° in the 6.27-7.42 GHz band, and the in-band |S 21 The difference is less than 0.1dB, the passband 3dB bandwidth reaches 1.15GHz (relative bandwidth 16.8%), and transmission zeros with depths of -33.13dB and -25.37dB are formed at the low frequency of 5.93GHz and the high frequency of 7.69GHz, respectively.

[0046] The layered structure design of the unit is as follows Figure 4 As shown, a six-layer metal stack-up architecture is adopted: the receiving layer is an H-shaped patch on an F4B dielectric substrate (εr=2.65, tanδ=0.0015), which is responsible for capturing electromagnetic energy; the modulation layer is an E-shaped open stub between the first reference ground layer (using Rogers RO4450F, εr=3.7) and the second reference ground layer (using RO4350B, εr=3.66), whose length controls the zero-point frequency and width adjusts the coupling strength; the transmitting layer is a U-shaped patch on an F4B substrate, which is coupled to the receiving layer through metallized vias; the DC feed layer uses a fan-shaped patch to achieve RF and DC isolation, and the specific dimensions are shown in Table I.

[0047] To optimize filtering performance, a three-step iterative design strategy is adopted. Figure 5 Cell1 has only symmetrical slots, with out-of-band suppression of only -1.83dB at 7.96GHz; Cell2 adds an E-shaped open stub, improving the suppression to -16dB at 7.84GHz; Cell3 adds a parasitic stub, forming a low-frequency zero of -33.41dB at 5.95GHz, ultimately achieving dual-zero filtering characteristics. Figure 6The performance of the initial unit, the unoptimized filter unit, and the optimized unit were compared. After optimization, the transmission zero point was shifted out of the passband, and the in-band transmission was improved by more than 5dB, thus solving the coupling interference problem between filtering and transmission.

[0048] Flexible zero-point shifting is achieved through parameter adjustment. Figure 7 Adjusting the length of the parasitic stub can independently control the position of the low-frequency null (5.93GHz), changing the parameters of the E-shaped stub can adjust the high-frequency null (7.69GHz), and adjusting the slot size can optimize the passband bandwidth without affecting the null, providing hardware support for dynamic spectrum adaptation.

[0049] An 8×8 array prototype was built based on cell design. Figure 8 It integrates 128 PIN diodes and connects to the FPGA control board via FPC flexible cables to achieve ±2.5V bias voltage switching. A darkroom lens testing system is built ( Figure 9 Four 8×8 sub-boards were arranged into a 16×16 array, and the S-parameters were measured using a Keysight E5071C vector network analyzer. The measured results are as follows: Figure 10 As shown, in State 1, the maximum transmission coefficient of the passband (6.24-7.36GHz) reaches -0.25dB, the low-frequency zero suppression depth of 5.72GHz is -40.67dB, the high-frequency zero suppression depth of 7.6GHz is -23.62dB, and the phase difference error is ≤±5°, which is in good agreement with the simulation.

[0050] Far-field pattern test, such as Figure 11 As shown, the distance between the transmitting antenna and the sample is 1.5m, the distance between the receiving antenna and the sample is 5m, and the center frequency is 6.8GHz. Figure 12 The measured results show that the array supports dual-beam scanning in a wide-angle domain of ±70°, and can form a stable beam at angles of ±18°, ±35°, ±55°, and ±70°, verifying the wide-angle control capability.

[0051] To verify actual communication performance, a QPSK communication system based on USRP-2974 was built. Figure 13 The transmit and receive antennas are spaced 0.8m apart, with an LF-TRIS embedded in the middle. Figure 14 Experimental results show that the video signal at the passband of 6.8 GHz is clearly restored and the constellation diagram converges; however, the signal at the out-of-band of 5.72 GHz and 7.6 GHz is severely distorted and the constellation diagram is chaotic, which confirms the practical value of strong out-of-band suppression.

[0052] This invention achieves integrated ultra-low loss, high-selectivity filtering, and wide-angle beam scanning through multi-layer coupling structure design, dual-zero point collaborative control, and multi-physics field optimization, providing hardware support for 6G anti-interference communication. Without departing from the core principles of this invention, adjustments can be made to the unit size and dielectric parameters; all such modifications fall within the scope of protection of this invention.

Claims

1. A low-loss filtered transmission type reconfigurable smart surface, characterized in that, The smart surface unit adopts a six-layer stacked structure, which includes, from top to bottom, the following: The receiving layer has symmetrical H-shaped metal patches etched on it, and two PIN diodes are symmetrically embedded in the hollow areas of the H-shaped metal patches; First reference stratum; The control layer has stub structures etched on it for controlling the zero point of high-frequency transmission; The DC feed layer adopts a fan-shaped metal patch structure and is electrically connected to the PIN diode of the receiving layer through a feed post; Second reference stratum; The emitter layer has slotted metal patches etched on it, and parasitic stubs for generating low-frequency transmission nulls are symmetrically arranged on both sides of it. The receiving layer and the transmitting layer are electromagnetically coupled through metallized vias that penetrate the six-layer stacked structure; 1-bit phase modulation is achieved by switching the bias states of the two PIN diodes; the stub structure of the modulation layer is coupled with the slot of the transmitting layer to generate a high-frequency transmission zero, and the parasitic stub of the transmitting layer generates a low-frequency transmission zero, thereby forming a bandpass filter characteristic with dual transmission zeros in the transmission frequency response.

2. The low-loss filtered transmission type reconfigurable smart surface according to claim 1, characterized in that, The branch structure of the control layer is an E-shaped open branch.

3. The low-loss filtered transmission type reconfigurable smart surface according to claim 1, characterized in that, The slotted metal patch of the emitter layer is a U-shaped metal patch, and the slot has a symmetrical structure.

4. The low-loss filtered transmission type reconfigurable smart surface according to any one of claims 1 to 3, characterized in that, The two PIN diodes include a first PIN diode and a second PIN diode; the 1-bit phase modulation is achieved by switching between two bias states: The first state is that the first PIN diode is reverse biased and the second PIN diode is forward biased; The second state is that the first PIN diode is forward biased and the second PIN diode is reverse biased.

5. The low-loss filtered transmission type reconfigurable smart surface according to claim 2, characterized in that, The position of the high-frequency transmission zero point corresponds to the length of the E-shaped open stub.

6. The low-loss filtered transmission type reconfigurable smart surface according to claim 3, characterized in that, The position of the low-frequency transmission zero point corresponds to the length of the parasitic branch.

7. The low-loss filtered transmission type reconfigurable smart surface according to claim 1, characterized in that, The receiving layer, transmitting layer and adjacent reference ground layer are bonded together using an F4B dielectric substrate; Rogers RO4450F material is used between the first reference ground layer and the control layer, and between the second reference ground layer and the DC feed layer; Rogers RO4350B material is used to bond the DC feed layer and the second reference ground layer.

8. The low-loss filtered transmission type reconfigurable smart surface according to claim 1, characterized in that, The metallized via forms a magnetic coupling path between the receiving layer and the transmitting layer; the series resistance of the PIN diode is 0.25Ω.

9. The low-loss filtered transmission type reconfigurable smart surface according to claim 1, characterized in that, The fan-shaped patch structure of the DC feed layer is used to isolate radio frequency signals from DC bias.

10. The low-loss filtered transmission type reconfigurable smart surface according to claim 1, characterized in that, The smart surface is composed of an array of multiple units, and each unit's PIN diode has an independent DC bias feed path.

Images