I. Introduction
The basic targets of ELINT are all types of radars, which are detected, located, and identified by their signatures in their operating modes (e.g., search and tracking). The signature (i.e., characteristics) of each radar consists of measurable parameters, such as the transmitter's frequency, power, mode, modulation, pulse width, and pulse repetition frequency (PRF). Thus, the signature of each radar is collected, analyzed, and stored. Special ELINT systems have been developed that scan each frequency band continuously, perform a real-time analysis of each intercepted signal, determine its signature, and compare this with others in a library to identify and locate the threat [1]. ELINT systems normally collect a lot of data over a long period of time to support detailed analysis [2], and ELINT determines what capabilities the enemy has, while radar electronic support (ES) determines which of the enemy's radars is being used at the moment and where the emitter (i.e., the weapon it controls) is located [3]. Generally, intercept receivers can be divided into radar warning receivers (RWR) and ELINT receivers. RWRs are designed to produce an immediate warning if specific threat signals are received. RWRs typically have poor sensitivity but work with a near-real-time processor that uses a few parameter measurements to identify a threat. Instantaneous frequency measurement (IFM) receivers are the most common among early-warning receivers. These types of receivers fulfill the fast-sweep requirements of the ELINT receiver [4].
The ELINT equipment, as shown in Fig. 1, is based on the superheterodyne (SH) receiver architecture but has a number of selectable intermediate frequency (IF) bandwidths. IF filters, coupled with a high-gain directive antenna, are able to achieve the very high sensitivity required to analyze signals that are also transmitted from the far lobes of the radar antenna. The selectable IF bandwidth is also required to analyze peculiar intra-pulse and inter-pulse signal modulations, which may provide individual identification (i.e., emitter fingerprinting) [5].
An electronic support measurement (ESM) for electronic intelligence applications places stringent demands on a receiver over a multi-octave bandwidth. Practical solutions must not only meet the electrical specifications, but also be compact and immune to microphonic effects and thermal fluctuations. A wideband receiver module was designed to meet these requirements and to provide matching phase and amplitude [6]. ELINT receivers must cope with a wide variety of signal parameters. The signals of interest extend over a wide dynamic range to handle different radar effective radiated power (ERP) values. The ELINT receiver equipment must cover a wide frequency range to search for new radars in any part of the spectrum. The modulation bandwidths of the signals can range from very short pulses of frequency modulation (FM) to continuous wave (CW) signals [7]. Search strategies using wideband receivers dedicate one of several receivers to the search function, determine the frequency of all signals present with wideband frequency measuring receivers, and perform a detailed analysis or monitoring with set-on receivers [8].
The concept for considering the operation scenario within the electronic support system was constructed so that the front-end modules were installed on the east, west, south, and north points to allocate 90° to each, enabling signals to be received across 360°. When a signal received in a specific direction is detected, only signals received from the corresponding path are precisely analyzed [9]. After that, the phase-matching path is switched to find a more accurate position. The design approach and contents proposed in this paper are described in the following paragraph.
First, the matrix module was assigned the function of selecting between the phase path and the amplitude path by using a transfer switch circuit, which consists of four switches. Second, a high-pass filter (HPF) was added between the output stage of the front-end (FE) module and the first stage of the matrix (MTX) module so that the cellular frequency did not flow through the antenna in the case of an unwanted signal (especially the mid-band, 2–6 GHz) coming from the outside. Third, since the amplitude path needs to receive the mid- and high-bands through the antenna at the same time, a power divider rather than a switch was used to monitor the amplitude in real time. At this time, an attenuator was added to each path to reduce the amplitude ripple for the frequencies of the different paths. Fourth, as several sub-modules must be connected with RF cables in the matrix module, structural design was performed from the disassembly/assembly point of view for performance improvement (debugging) after assembly [9]. Fifth, to secure the degree of isolation between the input and output channels, the mechanical structure was designed to prevent signals from flowing into other channels using valley form equipment [10].
II. Wideband Frequency Down-converter Structure
This paper describes an RF module that can receive 2–18 GHz and select a path from the front end of the rear end of the antenna to the matrix module. For each FE, there are 8 phase module (PM) inputs (2–6 GHz, 6–18 GHz), 4 outputs (2–18 GHz), and 1 BIT input (2–18 GHz) ports. The matrix has a very complex structure with 16 PM inputs, 8 amplitude module (AM) inputs, 8 output ports, and 4 BIT output ports. To protect the internal components in the case of an instantaneous increase in input power, a limiter is applied at the beginning of the FE module, and a low-noise amplifier is placed at the rear to secure an excellent noise figure. In addition, an HPF was applied to remove unwanted low-band frequency components for each path. Fig. 2 presents detailed configuration diagrams of the overall configuration of the four FE modules (east, west, south, and north) and the multi-channel matrix module required for the ELINT system. After low-noise amplification of the 2–18 GHz signal received from the antenna in the FE module, output of FE connected to input port of the matrix module through a 2-m RF cable.
III. Design and Simulation
1. Design of Transfer Switch to Select between Phase and Amp Path
In the ELINT system, from among the signals collected from all directions through free space, it is necessary to first identify roughly in which direction the signal is largely received. This can be confirmed through the amplitude-matching path. Then, the phase-matching path is switched to find a more accurate position. At this time, the function of selecting the amplitude-matching and phase matching paths in the matrix module is required. A circuit that performs these functions is defined as a transfer switch circuit. In order to select the corresponding path using four switches in the ultra-wide frequency band (2–18 GHz), the transmission line physically crossed. This could be solved by sending a signal. Figs. 3 and 4 present a configuration diagram and PCB layout.
2. Phase Matching for Adjustment using a Micro-Strip Line
After identifying the amount of signal received outside the RF front-end module, the FE and MTX modules located at the front end of the signal processing require phase matching between channels (4EA) to find the correct location after a specific direction (east, west, south, or north) is identified. At this point, a phase difference occurs for each channel, depending on the RF PCB, the ultra-high-frequency component, and the assembly state (especially the ground effect). Figs. 5 and 6 present an image and photo of an evaluation jig with a PCB layout drawing and an actual jig photo. Fig. 7 shows the maximum phase difference at 18 GHz as a measurement result. Based on the reference line, it can be seen that the phase delay is as Table 1. Table 1 shows the measured phase values at different lengths for the micro-strip line [11].
3. Isolation Design among RF Channels
In a module with multiple channels, the structure must be designed to prevent signals from flowing into other channels by using a mechanism to ensure that the isolation characteristics between input channels are excellent. It is also important to design a mechanism to prevent radiation caused by interference and crosstalk that can be generated in a very high-frequency band. To obtain the targeted isolation characteristics, a wall was mechanically formed along the RF path so that the signals that went over to free spaces other than the transmission line could be blocked [12] (Fig. 8).
4. Video Leakage
The RF front-end module must have both the antenna port that needs to receive the signal through free space and the BIT port required for the function to determine whether the internal circuit of the module is normal. In this case, the component required for path selection is the single-pole double-throw (SPDT) switch. However, in order for the switch to operate, it is necessary to select a path by receiving an external control signal.
In a typical PIN diode switch, the DC and RF signals share the same path, and a DC block is used to prevent the current from going into the load. When the control voltage changes states (low to high and vice versa), it actually charges the DC block capacitor and discharges through the external load, causing a video leakage signal to appear at the load. In FET switches, the drain and source are connected through the channel. FET switches have very low video leakage compared to PIN diode switches. The channel is controlled by the depletion layer of the gate bias [13]. Table 2 numerically shows how low the FET switch characteristics are compared to those of the PIN diode switch.
5. Analysis of the GAIN for Each Path
A broadband receiver module is vital to maintaining gain. To ensure good gain performance, a budget analysis tool should be used to identify its limitations. Typically, gain and noise figure (NF) represent trade-offs because they prevent two items from obtaining a good performance at the same time. Additionally, the characteristics of gain are important because they affect the signal thread level of an analog to digital (A/D) converter.
IV. Fabrication and Measurement
A GaAs MMIC chip was mounted on a gold-plated substrate to provide good grounding. A high-frequency Duroid PCB was placed around the chip to connect the DC bias, and the input and output lines were wire bonded using the chip-and-wire process [12]. The front-end receiver was made with 4 antenna (ANT) channels to select an input frequency between 2–6 GHz or 6–18 GHz, as shown in Fig. 10(a).
Fig. 10(b) shows an assembled image of the matrix box with three sub-modules (AM module, PM module, and DC/DC converter) and the RF semi-rigid cable. It can be seen that many cables are needed to connect the ports of the inner submodule to the interface of the outer side.
Fig. 11 shows a photo of a manufactured real module for three sub-modules using bare-type MMIC. Fig. 11(a)–10(c) shows pictures of a front-end receiver for the AM and PM, respectively: 8 CH ANT input and 4 CH output and 1 CH BIT, 2 CH ANT input and 1 CH output, and 4 CH input and 2 CH output and 1 port AM input.
1. GAIN and Return Loss Measurement Results for Each Module
The measurement data consist of the test results of the module alone. The gain in the band is the minimum value, and the reflection coefficient is the maximum value of the input port.
Fig. 12(a) for the FE module shows the gain graph and return loss, with a gain of at least 11.67 dB and a return loss of −7.21 dB at the maximum in the band (frequency 6–18 GHz). Fig. 11(b) for the FE module shows the gain graph and return loss, with a gain of at least 13.15 dB and a return loss of −16.89 dB at the maximum in the band (frequency 2–6 GHz). Fig. 12(c) for the amplitude path of the MTX module shows the gain graph and return loss, with a gain of at least 24.05 dB and a return loss of −6.55 dB at the maximum in the band (frequency 6–18 GHz). Fig. 12(d) for the amplitude path of the MTX module shows the gain graph and return loss, with a gain of at least 20.90 dB and a return loss of −7.82 dB at the maximum in the band (frequency 2–6 GHz). The measured results for each path are shown in Table 4, and these results are almost the same as the expected values, as shown in Table 3.
2. GAIN and Return Loss Measurement Results for Total Paths
This set of measurement data shows the results of the entire test connecting each module and the RF cable. The minimum value of gain in the band and the maximum value of the output port were measured for the reflection coefficient. Fig. 13(a) for the total path shows the gain graph and return loss, with a gain of at least 23.25 dB and a return loss of −11.63 dB at the maximum in the band (frequency 6–18 GHz). Fig. 13(b) for the total path shows the gain graph and return loss, with a gain of at least 22.99 dB and a return loss of −14.76 dB at maximum in the band (frequency 2–6 GHz). The measured results for each path are shown in Table 5. In a structure in which different modules are connected, impedance matching is very important to improve gain flatness and the reflection coefficient.
3. Phase Matching Measurement Results for Total Paths between Channels
This set of measurement data presents the phase differences between the channels (CH#2–CH#4). After normalizing channel #1, measured value shows the difference from the other channels. The peak values of phase matching among channels at 6–18 GHz were ±8.24°, as shown in Fig. 14(a). The peak values of phase matching among the channels at 2–6 GHz were ±3.30°, as shown in Fig. 14(b). The summary of the measured results for each band is shown in Table 6. As can be seen from Table 6, the measured phase-matching value is at a level that can more accurately identify the position of the enemy received through free space. In addition, and more important than the phase-matching value within the band, there should not be a sharp inflection point. This is because bad phase matching cannot follow linearity and cannot be overcome in the system calibration process. For phase matching between channels, a stable RF ground must be maintained to obtain excellent characteristics within a band without a discontinuous section. This means that, on the contrary, discontinuity occurs in a section with bad phase matching.
V. Conclusion
In this paper, we describe the design and fabrication of a FE receiver and MTX modules for 2–18 GHz that have several modules with high gain, good phase matching characteristics, and reliability; this was accomplished by applying a chip-and-wire process using a bare-type MMIC device. To compensate for the mismatch among many of the sub-modules, a FE module, MTX module, and BIT module suitable for sub-band frequency characteristics were designed and applied to the direct receiver. The broadband receiver module had two paths: a phase path and an amplitude path. Phase- and amplitude-matched RF semi-rigid cables of different lengths were used to connect to the internal sub-modules of the matrix receiver. The main RF line was a dielectric substrate, RT/Duroid 5880, with a relative dielectric constant of 2.2 and a dielectric thickness of 0.127 mm.
In the wideband frequency receiver module, the gain was 22.99 dB at mid-band (frequency 2–6 GHz), with a return loss of about 14.76 dB. The gain was 23.25 dB at a high band (frequency 6–18 GHz), having a return loss of about 11.63 dB. The peak values of phase matching among the channels for 2–6 GHz were ±3.30°, and the corresponding values among the channels for 6–18 GHz were ±8.24°. The proposed 2–18 GHz wideband frequency receiver module can thus be applied to a front-end amplifier that requires path selection at the back of the antenna of an ELINT system.