The S-Array glide path antenna system consists of three antennas, upper, middle and lower antenna. h₁ = 1.3h, h₂ = 2.0h and h₃ = 2.7h respectively. The structure of the antenna and the position of the near field monitor antenna can be found in Figure 1. Where h = λ 4 sin ( θ − F S L ) , θ is the glide angle, generally θ is 3˚, and Forward Slope (FSL) is forward slope in front of the antenna in first Fresnel zone. The frequency used below is 333.35 MHz and corresponding λ is 0.9 m. And h₁ = 5.6 m, h₂ = 8.6 m, h₃ = 11.6 m. Meanwhile the near field monitor (NF) is placedat a distance (D) about 57 m in front of the antenna system. The phase error between h₁ and h₃ to NF is 360˚. And the height is little higher than the value D ⋅ tan θ , mainly depends on tested value on the spot after the flight check.

The relative amplitude and phase for standard CSB, SBO, and Clearance (CLR) feeding of the three antennas are showed in Table 1. In antenna system distribution (ADU) the SBO amplitude is not related to CSB, the feeding coefficient k is changed by adjust SBO level, and the same to that of CLR.

The coverage range requirement can be found in Figure 2.

The lobbing diagram of three kinds of signal: CSB, SBO and CLR listed from 0 to 2θ are showed in Figure 3.

The space pattern of kathrein antenna is showed in Figure 4. From the side view of CSB coverage, to make up for lower angle coverage, the lobe is predominant by clearance instead of course. This design avoids reflecting from the far field obstacles effectively. Similarly, the SBO signal in lower angle (150 Hz predominant) own the lesser signal comparative to that of 90 Hz.

The theoretical course and clearance radiation patterns are described in Figure 5. The ddm function expression is:

ddm = 2 E SBO E CSB cos ( φ SBO − φ CSB ) .

The value of ddm is a constant value 0.4 for clearance signal predominant, while for course signal predominant, the value is twice the quotient of vector E_SBO divided by E_CSB. The boundary, the red cross-point, of them is marked in the top of the figure.

There are three requirements for ddm distribution around the glide angle θ.

Good linearity characteristic, symmetry and 75 μA (ddm = 0.0875) at 0.88θ. To get 75 μA, ddm = 0.0875 for lower half sector at 0.88θ, the coefficient k₁ for SBO relative level is:

7 160 ⋅ sin ( 1.3 h sin ( 0.88 θ − F S L ) ) − 0.88 sin ( 2 h sin ( 0.88 θ − F S L ) ) + 0.14 sin ( 2.7 h sin ( 0.88 θ − F S L ) ) − 0.535 sin ( 1.3 h sin ( 0.88 θ − F S L ) ) + sin ( 2 h sin ( 0.88 θ − F S L ) ) − 0.535 sin ( 2.7 h sin ( 0.88 θ − F S L ) )

where the h is electric length: h = π 2 sin ( θ − F S L )

ddm = −0.0875 for upper half sector at 1.22θ, the coefficient k₂ is:

− 7 160 ⋅ sin ( 1.3 h sin ( 1.22 θ − F S L ) ) − 0.88 sin ( 2 h sin ( 1.22 θ − F S L ) ) + 0.14 sin ( 2.7 h sin ( 1.22 θ − F S L ) ) − 0.535 sin ( 1.3 h sin ( 1.22 θ − F S L ) ) + sin ( 2 h sin ( 1.22 θ − F S L ) ) − 0.535 sin ( 2.7 h sin ( 1.22 θ − F S L ) )

Approximately, k₁ = k₂ = 0.173. The further calculation results with different FSL from −0.5˚to 0.5˚ are listed in Table 2. The data show that the coefficient k₁ is decreased with the increasing of FSL. And the error of k₁ and k₂ is negligible, which demonstrates the good symmetry. Discussion about ddm curves below would further demonstrate the linear variation around θ.

Three curves of ddm distribution are exampled in Figure 6, flat ground and the ground with a forward slope for FSL = ±0.5˚. It can find that the course ddm shorted while the FSL rise, meanwhile the feeding coefficient increase. Seen in green curve, the linearity fades to be destroyed when the elevation angle surpass 4˚. But for other two curves, both of them possess better linearity. In half glide path section region, ddm from −0.0875 to 0.0875, all of three curves are perfectly coincident. This shows the good linear behavior and good symmetry.

The antenna elements should be aligned along a straight line, it shall be perpendicular to the average forward slope (FSL). If the forward slope is rising, the lower antenna shall be forward compared to the middle one.

The sideways offset of the antenna elements shall be accurately adjusted. Orientation is such that the upper antenna is closer to the runway than the middle one. And the middle antenna shall be closer to the runway than the lower one.

Detailed, for the formula below, FSL positive if ground rising toward threshold, SSL positive if ground plane rising from antenna mast toward runway.

The formula of the alignment parameters are listed below:

The longitudinal offset lower to upper (sft_d) is: 1.4 h ⋅ sin ( F S L ) , middle to upper (sft_m) is 0.7 h ⋅ sin ( F S L ) . And lateral offset upper to middle (off_u) is:

k₁ for 0.88θ, k₂ for 1.12θ and Δk = k₁ − k₂.

( 2 2 − 2.7 2 ) ⋅ h 2 2 ⋅ D + h ⋅ sin ( S S L ) ,

lower to middle (off_d) is:

( 2 2 − 1.3 2 ) ⋅ h 2 2 ⋅ D + h ⋅ sin ( S S L ) .

where h is: h = λ 4 sin ( θ − F S L ) .

It is worth noting that the Side Slope (SSL) just affects sft only, but FSL would affect the height of antenna, off and sft.

The case of FLS = 0 and D = 120 m: h₁ = 5.58 m, h₂ =8.59 m, h₃ = 11.60 m, off_u = −25.3 cm, off_d = 17.7 cm.

The phase of the total received signal at the monitor point M will be as shown in Figure 7. Where Φ_M = (Φ_r + Φ_d)/2, and Φ_r is the electrical distance from radiated antenna to the receive point of the direct signal, similarly Φ_d works as the electrical distance of the reflected signal.

To monitor the glide path angle, the phase error should be 0˚ or 360˚. For a nearly 0˚ phase error, the distance must be more than one kilometer. This distance corresponds to a very high monitor antenna. But for 360˚ phase error, the distance is approximately 57 m. The phase error Φ_MU − Φ_ML for θ = 3˚ as a function of the distance L in front of the antenna system is shown in Figure 8. The former three kinds of glide path antenna, null reference (NR), sideband reference (SR) and M-Array, are also listed for comparison. The type of NR and SR are possessed of two antennas, neither 0˚ nor 360˚ are convenient, for a nearly 0˚ phase error the distance must be more than 500 m, while for 360˚ phase error the distance is 30 m for NR and 21 m for SB. This very short distance results in a 16˚ elevation angle for NR and 20˚ for SR to ground reflected signal from the upper antenna element. So the monitor position chosen for 180˚ is advisable, the

distance is approximately 62 m to capture the zero point of ddm for NR. In particular, the monitor antenna located at about 41 m and 2θ (6˚) to obtain the zero point of ddm for SR. Similar to S-Array, the distance is approximately 82 m for M-Array. Above all, besides the distance for SR (~41 m), the S-Array one be possessed of a shortest distance of the currentimage type glide path antenna. It means the height of monitor antenna is also the lowest among four types. These characteristic is advantage to install and maintain, at the same time, reduce the cost.

For near field monitor, the ddm curve is essential to study. This is the monitoring basis from which the far field space signal distribution could be speculated. So the ddm distribution at 57 m (Φ_MU − Φ_ML = 360˚) from the antenna system is given in Figure 9. Around the monitor point (ddm = 0) the symmetry of the distribution becomes decayed. Specifically, the lower and upper half sections (ddm = ±0.0875) are 2.887˚ and 3.164˚ respectively. In fact, the ddm distribution has been disrupted. It can be found that the second glide path appears at approximate 3.6˚. This is the false glide path, among 3˚ - 3.6˚ 150 Hz predominant, besides 90 Hz predominant from 1.5˚ to 4.5˚ region.

Despite the alignment of lateral offset, the ddm curve would remain disrupted in NF, especially near the threshold. The direct result is the deviation of the zero point of ddm. In approaching range, the ddm distribution from 600 m to the threshold is presented in Figure 10. In order to have a good vision, the ddm is the absolute value, and the write line at bottom represents the line of ddm = 0. Clearly, with the distance closes to threshold, the ddm curve moves the upper angle, also the zero point of ddm follows to the high elevation angle. Finally, this leads to the glide path upturned.

To research the NF condition further, the characteristic of ddm not only on glide path, but also in front of the NF are also studied. The angle (ddm = 0) of two cases versus corresponding distance are showed in Figure 11, the inset gives the theoretical height versus the distance. For the case on glide path (case one), the red line in Figure 11, also the write line in Figure 10, indicates the angle is about 3.2˚ at threshold, a litter higher than the glide path angle (3˚) in far field. However the case of signal in front of the near field monitor (case two) is completely different. Firstly, the angle gradually becomes to lower when closes to the threshold. Secondly, there is a sharp decline within the distance of 1200 m (closes to the mark of NF: 1160 m), declines to 1.8˚ at threshold. This variation is 6 times (Δ θ = 1.2˚) than that on the glide path. Compare to the ddm distribution at near field monitor (~57 m) position, the ddm distribution in 0 - 600 m is better, as least there is not false glide path, and the signal in their respective dominant area is right. The reason why the ddm distribution in case one is better than case two contributes to the lateral distance and it’s corresponding lateral offset. To large extant, this alignment offsets the phase error both in FF and NF. The original purpose of the siting deviated from the one side of the runway is avoid touching aircraft. Lucky, this deviated siting also could improve near field signal by lateral offset.

After the glide path iron tower has been erected, it is not easy to move the position

of each antenna, especially the old one which has been used for years. During the flight check, sometimes the position of the antenna should be aligned, such as mending glide angle by align the height of the middle antenna, optimizing the structure of III zone by adjusting offset, and improving TCH by twisting the antenna and so on. It is essential to change to the antenna position as soon as possible. In other words, the less extent, the quicker finished. The extent of several offsets of two types is listed in Table 3.

Theoretical calculation demonstrates that it is more easily to improving TCH for S-type than M-type by rotating the upper antenna. Obviously the S-type has the less changing extent. The parameter h₃ is the symbol of the size of the antenna system, the larger it is, the more expensive and inconvenient to maintain. The parameter h₃ − h₁ represents the working range to maintain or repairmen. For example, if the transmitter cable is damaged, the ADU should be calibrated again after replaced the new cable, and operator would connect the cable up and down in this working region, the lager this region the poorer efficiency. And so do lateral offset adjustment.

The ddm distribution for near field monitor has been listed in Figure 9. Although the poor symmetry of the curve, it has both upper half width (ddm = −0.0875) and lower half width (ddm = 0.0875). By contrast M-type just has the lower half width. And the linearity of S-type around the θ is more excellent than that of M-type.

For the monitor antenna, S-type is lower and has a shorter distance to glide path tower. This means the lower cost for manufacture and convenient to adjust monitor to the position of ddm = 0.

The beam bend potential (BBP) represents the anti-interference ability, the less bbp value the better it works. S-Array has the lowest radiation in lower angles to

Note: The case of θ = 3˚, f = 333.35 MHz, FLS = 0.3 and D =120 m. ΔM means the variation from FLS = 0.3 to FLS = 0.1, the other parameters reserved. All the units of the data above are meter.

reduce illumination of terrain objects in order to reduce bends. Strictly speaking, the radiation of M-type is a little lower than S-type below 0.7˚, because the S-type appears negative value, but after that, it owns the lower radiation up to 3˚. As seen in Table 4, the S-type has the lowest bbp value at 1˚, far less than the other three. This illustrates the good anti-interference ability for S-Array. Meanwhile the coverage parameter, CSB, chosen the value at 1˚ divided by its maximum within 6˚, shows that the S-type also has the minimum value. This confirms the satisfactory performance of low angle obstacle depressed (Figure 12).