Active antenna modules can offer a variety of benefits for aircraft avionics systems. These modules provide directional and omnidirectional modes, transmit and receive modes, and direction-finding (DF) capability with acceptable bearing accuracy. In addition, the use of active circuitry can result in dramatic decreases in the size and weight of avionics systems through the careful selection of such components as amplifiers and cables.

The last decade has been marked by rapid developments in integrated avionics and other systems.^1-3 Integration brings with it reductions in size, weight, cost, and power consumption, along with useful features such as built-in-test-equipment (BITE) capabilities.

As an example, conventional Traffic Collision and Avoidance System (TCAS) and Transponder systems employ four separate antennas, ten cables, and separate receivers and transmitters. In contrast, an integrated TCAS/transponder system has two combined antenna modules,^1-3 electrically connected to the single TCAS/transponder transmit/receive unit.

Active antenna modules can switch operating modes as needed. The directional antenna mode should be implemented for a TCAS 1090-MHz receiver, TCAS 1030-MHz transmitter, and Transponder 1030-MHz receiver functions. The omnidirectional antenna mode should be implemented for Transponder 1090-MHz transmit and, in some instances, TCAS 1030-MHz transmit functions.

This article will review the use of active antenna modules for avionics integrated systems — including TCAS amplitude monopulse systems, which require directional antenna mode. Other systems that can be served by active antenna modules include Transponder, Universal Access Transceiver (UAT), Automatic Dependent Surveillance-Broadcast (ADS-B), and Distance Measurement Equipment (DME) systems, which require an omnidirectional antenna mode of operation. Table 1 shows active antenna module specifications for different L-band systems.

The common TCAS/Transponder antenna module³ can provide directional (or, if necessary, omnidirectional) TCAS antenna radiation patterns, omnidirectional Transponder radiation antenna patterns, and directional TCAS and Transponder receive antenna patterns. The TCAS directional antenna is a four-monopole, vertically polarized array capable of transmitting in four selectable directions at 1030 MHz.^3,4 TCAS direction-finding (DF) systems rely on the amplitude-comparison method or the phase comparison method.

This report will consider the first of these techniques, which is relatively simple and cost effective, but relatively inaccurate. The antenna of this system is capable of receiving replies from all directions simultaneously with bearing information at 1090 MHz, using amplitude-ratio monopulse techniques.⁵ The TCAS system uses signals from the directional antenna to determine the bearing from the host to another aircraft. The novel antenna module consists of the four-monopole antenna, switched beam forming network (SBFN), matching network, and interface.⁴ Each SBFN port is coupled through a cable to the corresponding receiver.

The approach used by airborne amplitude monopulse system is to estimate an intruder aircraft's bearing by comparing magnitudes of signals received by the four-monopole directional antenna. The antenna aperture is divided into four sectors (or quadrants). The incoming signals are processed inside the antenna module to produce four electrical connector signals, so that each electrical signal represents a unique quadrant of the polar coordinate system.

A significant bearing accuracy improvement can be achieved with the bearing algorithm (or index) which uses all four received signals in the four sectors. In this case, the bearing may be detected not only by the main lobe(s) of the antenna pattern, but also by the side or back lobe(s) of another beam(s). This additional information provides greater bearing accuracy and eliminates strong requirements for the side or back lobe level suppression.

Generally, L-band TCAS, Transponder, UAT, and DME avionics systems employ separate antennas, cables, receivers, and transmitters. Thus, each aircraft may include top and bottom TCAS antennas, top and bottom Transponder antennas, top and bottom UAT antennas, and a bottom DME antenna. Separate cables are required for each antenna. Using separate systems with passive antennas adds to the complexity, cost, size, and weight of the system and typically results in poor performance.

Conventional separate TCAS antenna arrays are passive systems. The existing TCAS transmitter with about 200 W pulse power provides a transmit range of 45 nautical miles given the typical directional antenna gain. During the receive operation, signals that are received by the antenna elements are conveyed at very low power levels over the transmission paths to the transceiver, without any power boosting from the passive antenna array.

The system also has limitations in the receive range and receive sensitivity (MTL) due to losses and noise produced by the coaxial transmission line (cable) and the associated circuit components at the receiver and the transmitter. For example, a conventional system may only be able to provide receive sensitivity for up to 90 nautical miles. The limited receive sensitivity is due in part to cable losses, in part to noise within the receive signal path and the like. In certain applications, it is desirable to provide a TCAS receive range of 100 nautical miles, which may not be possible with these conventional passive antenna systems.

Another disadvantage of the existing systems is the influence of the mismatched antenna interface (cables, transmit/receive block, and connectors), which is detrimental to the antenna pattern shape, bearing accuracy, and antenna gain. The beam-forming network (BFN) of the passive antenna module is very sensitive to the mismatching of the antenna interface.

A typical avionics antenna BFN includes four two-branch hybrids.^4-6 Table 2 offers values for isolation and phase error of a two-branch hybrid and parameter variations of a passive antenna module as a function of the interface return loss at 1090 MHz. The deviation of the interface return loss from 16.0 to 10.0 dB causes degradation of the antenna performance.

In conventional systems, during the TCAS directional transmit mode, the transmit channels (including cables) are alternately activated. According to ARINC requirements,⁷ the maximum cable loss variance is 1.0 dB. Unpredictable insertion loss variance causes an imbalance between the channels. Therefore, to provide an identical antenna pattern performance from sector to sector (for four directions) during the transmit mode, an amplitude calibration of transmit signals is required. To improve the TCAS bearing accuracy of an amplitude monopulse system, both an amplitude calibration of the receive channels and antenna lookup tables (LUTs) are used.⁵

These LUTs are usually prepared with a well-matched antenna interface (test equipments). But in real life, the mismatched antenna interface (due to cable performance variance, environment conditions, manufacture tolerances, etc.) causes additional bearing errors. The above disadvantages can be eliminated by using the active antenna described in this report.

The key requirements for active antennas to be used in avionics systems are good electrical performance, compactness, light weight, low cost, low DC power consumption, and high reliability. Active antennas may be classified into three basic categories: transmitting, receiving, and both transmitting and receiving. An active receiving antenna is the one that contains some electronic circuitry, which can amplify a received signal at the antenna and thus avoid interference that may enter the system at the transmission line (cable).

Another purpose of an active antenna is to transform unusual antenna terminal impedance into a constant value that matches the characteristic impedance of the cable. An active antenna usually costs more than a passive antenna and consumes more power, but its electrical performance is typically much better.

Figure 1 shows a block diagram of an active antenna module where four antenna monopoles — A1, A2, A3, and A4 — are connected to a nonswitchable BFN [see Figs.2(a) and 2(b)] or a switchable BFN (SBFN) [Fig. 2(c)]. The SBFN includes a BFN and a switched 0/180-deg. phase shifter.^3-5 A transmit signal passes through one antenna module port 24, single-pole, double-pole (SPDT) switch SW5, power amplifier (PA), and SPDT switch SW4, then splits and switches between inputs 1, 2, 3, and 4 of the antenna BFN (or SBFN).

During the directional transmit mode, each BFN input is activated alternately to provide the directional antenna pattern for one quadrant (forward, right, aft, or left). Therefore, the position of the directional pattern during the transmit mode depends on which BFN terminal is activated. In this case, only one of the four radiation lobes is present.

Each of the four antenna module inputs corresponds to a beam in one of four directions: front, right, aft, or left. For the SBFN the directional transmit mode is implemented by the alternate activation of one of the input ports (1, 2, 3, or 4), while the switched phase shifter provides 0-deg phase shift. During the omnidirectional transmit mode in the system with nonswitchable BFN [Figs. 2(a) and 2(b)], all four BFN inputs (1, 2, 3, and 4) are activated simultaneously by four signals with equal amplitudes and phases.

For the network with the SBFN [Fig. 2(c)], during the omnidirectional transmit mode, the transmit signal passes only through input 2 of the SBFN while the switched phase shifter is in the 180-deg. phase-shift state.⁴ The antenna provides the omnidirectional pattern because the four antenna monopoles (with quarter-wavelength space between the diagonal elements) are activated with equal magnitudes and progressive 90-deg. phase shifts (0, 90, 180, and 270 deg.).

During the receive directional mode, all four antenna module terminals (23, 24, 25, and 26) are monitored. Received signals from the BFN pass through LPFs, switches, LNAs, and cables to the transmit-receive block. In the one common receive-transmit channel (see ^{Fig. 1}) of the antenna module, the received signal passes through the four-way splitter, two SPDT switches, and low-noise-amplifier (LNA) LNA2. The incoming signals are processed inside the active antenna module in each channel to produce four electrical signals, such that each electrical signal represents a unique quadrant of the polar coordinate system.

The relative signal intensity from four receivers shows the azimuth direction of a selective object according to the special bearing algorithm.⁵ In this case, the bearing may be detected not only by the main lobe(s) of the antenna pattern, but also by the side or back lobe(s) of another beam(s).

The performance of different quadrature hybrids for a possible antenna BFN application was considered in Ref. 4. Two-branch hybrids have low insertion loss, low phase error, and adjacent output ports which permit combining them in the planar BFN design. The direct connection of the four two-branch hybrids (without additional connection lines between them) makes the bandwidth of the 4 × 4 matrix slightly narrower than the bandwidth of the single two-branch hybrid due to the undesirable interaction between the four hybrids.

When the four hybrids H1, H2, H3, and H3 are connected using quarter-wavelength transmission lines (invertors I1, I2, I3, and I4) [Figs. 2(a) and 2(b)], the quality of the circuit is improved. The bandwidth of all S-parameters is considerably widened and the return loss and isolations are better than 20 dB in about a 20% bandwidth.⁶

The two-branch hybrids can be used at L-band BFN [Fig. 2(a)] for relative narrowband systems [TCAS, TCAS/Transponder, TCAS/Transponder/UAT, TCAS/ADS-B) (Table 1)]. The three-branch hybrid has a wider bandwidth than the two-branch hybrid but a greater insertion loss and greater dimensions.

In the broadband coupled line hybrids,⁶ coupling tolerances are difficult to realize because of the narrow gap between the strips. However, for the antenna modules in the combined TCAS/Transponder/DME system with bandwidth at 30%, the coupled line hybrids [Fig. 2(b)] should be used.

Active devices in the antenna module always present undesired higher-order harmonics, which should be suppressed through a properly designed LPF. The four BFN terminals are connected (Fig. 1) to lowpass filters (LPFs). The latter remove any high-frequency harmonics that may appear in the transmit signal and protect LNAs and front-end receivers from unwanted high frequency receive signals from outside sources.

Also, LPFs reject high parasitic harmonics from non-linear elements, for example switch PIN diodes. LPFs are also used for impedance matching between BFN terminals and the active antenna switch network. Different configurations of print LPFs and their performance are shown in Fig. 3 and Table 3, respectively.

The implementation of LNA in an active antenna structure increases antenna gain and bandwidth, while improving noise performance. An active antenna module with the LNA provides a greater receive sensitivity because the lossy cables and the transmit-receive block after the LNAs have a negligible effect on the overall receiver input noise figure (NF).

The NF contribution of the transmission cables can be effectively eliminated. This can be as much as 3 dB,⁷ which is a substantial portion of the receiver's noise budget. Also, the mismatching of the antenna module interface (cables, transmit-receive devices) has a negligible effect on the antenna module performance: bandwidth, insertion loss, switching speed, isolation, return loss, RF power, etc.

The design of the LNA has to achieve the following main goals: good antenna return loss (RL), low NF, stability, and high gain (G). The design targets for the LNA of the active antenna module are noise figure NF < 0.5 dB, gain (G) >15.0 dB, return loss (RL) < 15 dB, low-power supply, and minimum number of components. To satisfy these requirements, the MGA-61563 LNA integrated circuit (IC) from Avago Technologies can be used.¹¹ Usually, LNA design is a tradeoff between input NF and input matching. The input matching of an LNA is the most critical parameter because it influences the performance of the BFN and the antenna module (see Table 2).

In modern aircraft, space, weight, and cost are of high importance. In the TCAS and combined avionics systems including TCAS, the eight RF cables (four for connection with the top antenna and four for connection with the bottom antenna) substantially increase the weight of the electronics network. For the passive antenna module, all eight identical cables provide transmitting and receiving of RF signals. In these systems, excessive cable losses degrade both transmitter output power and receiver sensitivity. The cables with lowest possible losses have a larger diameter but their weight, space, and cost are relatively high.

In a system with active antenna modules, only one cable [cable 2 (Fig. 1)] and one bottom cable (not shown in Fig. 1) support both transmit and receive functions, while six other cables (the three top and three bottom cables) are used for receive only. Smaller diameters can be used for these six receive cables to minimize weight, size, and cost. Excessive loss in these six cables is not critical in to receiver sensitivity because LNAs are placed before the cables (Fig. 1). This positioning of the LNAs yields a better cumulative receiver NF than that with the passive antenna module structure.

For example, transmit/receive cable 2 can be RG 214 with an outside diameter of 0.405 in., which exhibits loss of 8.0 dB/100 ft, but weight of 0.11 lb/ft and cost of $83.00/100 ft. Including the 0.25-dB loss of the connector at each end of the transmit/receive cable assembly leaves an allowance of 1.5 dB maximum cable loss according to ARINC requirements⁷ which also provides minimum transmit loss.

The other cables (1, 3, and 4), used for the receive mode only, can be smaller-diameter cable, such as RG 58, with an outside diameter of 0.195 in., with loss of 23.0 dB/100 ft, weight of 0.026 ft/100 ft, and cost of $45.00/100 ft. By optimizing the cables, the active antenna module provides a substantial reduction in weight and cost for an avionics system.

The active antenna module (Fig. 1) includes five SPDT switches and one SPST switch. Design of these switches depends on their electrical requirements: bandwidth, insertion loss, isolation, RF power, etc. The conventional SPDT switch, which includes quarter wavelength lines, is too large for the L-band. Figure 4(a) shows a microminiature SPDT switch based on the dumbbell-shaped defected ground plane DGS.^9,12

The dumbbell-shaped DGS [Fig.4 (c)] includes two wide defected areas connected by a narrow slot. The slow-wave effect is utilized to reduce the circuit size of the RF circuitry. Insertion losses of the conventional quarter wavelength microstrip line [Fig. 4(b)] and the DGS microstrip line [Fig. 4(c)] are identical.

The DGS microstrip switch network [Fig. 4(a)] provides a 50% size reduction as compared to the quarter-wavelength line switch. In order to protect the receiver from high power during the transmit mode, a high isolation of the receiver path is required, but in the receiver mode, a high isolation of the transmit path is not necessary.

Therefore, in the TX network [see Fig. 4(a)], only one PIN diode (instead of two) can be used to provide less TX loss, a cost reduction, and a smaller area of the switch PCB. In this case, 20 dB of isolation in the transmit arm is sufficient to prevent any variation in the output impedance of the transmitter from affecting the performance of the receiver.

In the receiver arm with two PIN diodes, more than 40 dB of isolation is necessary to protect the receiver from the strong transmit signal. High-power surface-mount PIN diodes, such as model 4P505-1072T diodes from M/A-COM Technologies (www.macomtech.com) or model GC42405-M1 from Microsemi (www.microsemi.com) can be used. The bias of the PIN diodes can be controlled from the transceiver block in any appropriate manner, such as by using the central conductors of two of the antenna cables.

The four-way splitter in the active antenna module of Fig. 1 can be implemented using the existing surface mount splitters, for example a model SCA-4-20 from Mini-Circuits (www.minicircuits.com) or model ES-4-7-V2 from M/A-COM Technologies. However, the majority of the existing splitters do not satisfy the requirements of the broadband active antenna module: bandwidth greater than 30%, insertion loss of less than 1.0 dB, isolation of greater than 15 dB, and power-handling capability of several hundred watts pulsed power. To resolve this problem, a combination of two-way and three-way splitters can be used.

The active antenna (Fig. 1) includes a power amplifier (PA) for amplification of a transmit signal. In conventional systems, the PA is part of the transmitter which is separated from the passive antenna by coaxial RF cables. For these systems, the radiated power level depends on the cable insertion loss. According to the ARINC requirements,⁷ the maximum cable loss is 3.0 dB. Such a loss will reduce the output power of the transmitter.

To eliminate this disadvantage, the final PA is placed inside the active antenna module. The PA can amplify the transmission signal in any conventional manner to a power sufficient to drive antenna according the MOPS requirements.^13-16 The latest laterally diffused metal oxide semiconductor (LDMOS) transistor technology is capable of a single device with 200-W output power over the 1030-to-1090-MHz frequency band using an all-gold metallization system.

In the active antenna module with a transceiver, the high isolation SPDT SW4 and SW5 are used to achieve isolation greater than 45 dB between transmit and receive channels. The other version of the active receive mode can only be implemented without the PA inside the antenna module. In this case, the passive bypass transmission line 18 - 19 (Fig. 1) allows a transmit signal to pass without amplification.²

To improve bearing accuracy in the amplitude monopulse system, the amplitude calibration of all four receiver channels should be implemented.⁵ The amplitude calibration provides protection against receiver channel imbalance, cable loss variation,⁷ aircraft elevation angle variance^13-16, temperature extremes, manufacture tolerances, etc. There are two possible calibration procedures: (1) the calibration signal is routed through the receiver channel only, and (2) the calibration signal is routed both through the antenna and through the receive channel (complete calibration).

In the first calibration procedure, the calibration signal passes to the common transmit/receive channel (through cable 2 in Fig. 1) via two SPDT switches SW5 and SW4 through the bypass line and the four-way splitter. During the first step of this calibration procedure, the calibration signal passes to three receiver channels through the three SPDT switches SW1, SW2, and SW3 to provide passing of the three equal calibration signals in the three receivers LNA1, LNA3, and LNA4.

These calibration signals are occurring simultaneously and, after passing through RCV1, RCV3, and RCV4, are memorized in the FPGA (not shown in Fig. 1). The calibration signal does not pass to terminal 2 because the SPST SW is in the OFF state. The second calibration step includes the calibration of the transmit receive channel when the calibration signal passes through the SPDT SW5, bypass line, SPDT SW4, LNA2 while the SPDT SW1, SPDT SW3, SPDT SW4, and SPST SW are OFF for the calibration signals.

After amplification by the LNA2 with an approximate gain of 20 dB, the calibration signal passes through the SPDT SW5 which has approximately the same 20 dB isolation. As a result of the equal LNA gain and switch isolation, the calibration signal level in the transmit/receive channel is equal to the original calibration signal. This signal passing through RCV2 is memorized in the FPGA and compared to the three calibration signals in the previous step 1. Finally, the four calibration signals passing through the four receivers are equalized.

The second possible calibration procedure includes the BFN and the antenna. The calibration signal from the four-way divider alternately activates each BFN terminal 9, 10, 11, and 12 through the SPDT SW1, SPST SW, SPDT SW2, and SPDT SW3. For example, during the passing of the calibration signal to the BFN port 9, the other ports 10, 11, and 12 are not activated because the SPST SW, SPDT SW2, and SPDT SW3 are closed for the calibration signal.

As a result of activating terminal⁹, all four antenna monopoles A1, A2, A3, and A4 are activated through the BFN. Coupling between antenna terminals is used in this calibration procedure. Coupling between the antenna inputs/outputs depends on the nonideal isolation between inputs/outputs of the SBFN hybrids and the mutual coupling between the four folded antenna monopoles.⁴ The isolation in the hybrid BFN differs from the ideal due to antenna interface mismatching, losses in the transmission line, amplitude and phase imbalance of the two-branch hybrids, hybrid discontinuities, channel imbalances, differences in insertion amplitude and phase between the four channels (interface), and BFN manufacturing tolerances.

The mutual coupling dominates in the small antenna array where the folded monopoles are closely spaced. Due to mutual coupling between the antenna monopoles and leakage between the BFN terminals, the calibration signals pass through the SPST SW2 and SPDT SW3 (which are opened for receive signals) to the RCV3 and RCV4 and memorized in the FPGA. A similar procedure should be implemented when the other RCVs are calibrated. Finally, the complete calibration including both the BFN and the antenna is implemented.

The size and weight of the active antenna module depend on electrical performance, transmission line type, dielectric substrate parameters, component dimensions, and structure of the network which can be implemented as a multilayer design, a three dimensional design, or a vertical-horizontal design.⁶

The main objective of the multilayer PCB is to increase significantly the density of the active antenna module. The benefits of this design can include size and weight reduction, enhanced performance, improved reliability, and decreased system cost. Multilayer design incorporates RF, digital, and control functions. Interconnections in a multilayer PCB include vertical transmission lines and/or so-called vias.

The most contradictory parameters of the active antenna module are: cost versus size, cost versus insertion loss, and insertion loss versus size. These parameters should be optimized by a tradeoff design. The major considerations in the “cost-versus-loss” tradeoff design are the types of transmission line and dielectric substrate. The combination of a microstrip line with a DGS and a stripline offers certain advantages of both cost and insertion loss. The substrate thickness also causes a contradiction between cost and loss.

The positive effects of decreasing substrate thickness are lower cost of materials and lower RF circuitry size. However, a decrease in the substrate thickness leads to a higher microstrip conductor loss. For the active antenna module the optimum substrate thickness is 10.0 to 20.0 mil. Table 4 shows the performance of the active antenna with the SBFN for 1030 to 1090 MHz.

1. L. G. Maloratsky, “RF Design of Avionics L-Band Integrated Systems,” Microwave Journal, October 2009, pp. 64-82.

2. D. Kutman et al., “Multifunctional Aircraft Transponder,” United States Patent No. 6,222,480, April, 2001.

3. L. G. Maloratsky et al., “Combined Aircraft TCAS/Transponder with Common Antenna System,” United States Patent No. 7,436,350, October 14, 2008.

4. L. G. Maloratsky, “Switched Directional/Omnidirectional Antenna Module for Amplitude Monopulse Systems,” IEEE Antennas and Propagation Magazine, November 2009.

5. L. G. Maloratsky, “Analyze Bearing Accuracy of a Monopulse System,” Microwaves & RF, Part 1, March 2009, Part 2, April 2009.

6. L. G. Maloratsky, Passive RF & Microwave Integrated Circuits, Elsevier, Amsterdam, 2004.

7. ARINC Characteristics 735, 1993.

8. Xiao-ming Zhong, “Design of Low-pass Filter Based on a Novel Defected Structure,” Journal of Shanghai University, Vol. 11, No. 4, August 2007.

9. L. G. Maloratsky, “Microstrip Circuits with Modified Ground Plane,” High-Frequency Electronics, December 2009.

10. D. Kurita et al., “Super UWB Lowpass Filter Using Open-Circuited Radial Stubs,” IEICE Electronics Express, Vol. 4, No. 7, 2007, pp. 211-215.

11. Avago Technologies, data sheet for model MGA-61563 current-adjustable low-noise amplifier.

12. Kim Dong-Wook, “Small-Sized High Power PIN Diode Switch with Defected Ground Structure for Wireless Broadband Internet,” ETRI Journal, Vol. 1, February 2006, pp. 84-86.

13. ARTSRBS/Mode S MOPS, RTCA/DO-181C, 2001.

14. DME MOPS, RTCA/DO-189, Inc. 1985.

15. UAT MOPS/DO-282, RTCA, Inc. 2002.

16. ADS-B MOPS/DO-260A, 2003.