Radar Losses
5.3 ANTENNA LOSSES
Fig ure 5.3 Example T/R module block diagram.
As an example of total transmit RF losses, Table 5.2 contains a summary of these losses for the three transmitter configurations of Figures 5.1 and 5.3. In computing the waveguide losses, we will assume the radar with the reflector (top drawing of Figure 5.1) is an L-band search radar. The space-fed phased array in the lower drawing is an S-band multifunction radar and the T/R module in Figure 5.3 is used in an X-band multifunction radar. The difference in operating frequencies is the reason for choosing the different waveguide losses.
Table 5.2
Example Transmit RF Losses
5.3. The entries for waveguide and stripline feed apply to antennas that use constrained feeds, and the difference between parallel and series feed networks is illustrated in Figure 5.4 [22]. In the series feed, the energy enters on one end of an RF transmission line (such as a rectangular waveguide, stripline, or microstrip) and is extracted at different points along the line. In a parallel feed network, the energy enters an RF transmission line and is subsequently split several times before being delivered to the radiating elements. As a note, it is possible for an antenna to use both series and parallel feed networks [23, pp. 5–8]. As an example, the rows the array could be fed by a series feed, while the elements in each row would be fed by a parallel feed network. It will be noted that the feed loss assigned to active arrays is 0 dB. This is because the radiating element driven by a T/R module is very close to the power amplifier.
The phase shifter losses apply to passive and constrained feed phased arrays. As a note, the losses apply to the entire array and not to each phase shifter of the array. The losses are shown as 0 dB for active phased arrays because the phase shifter is not in the path between the antenna and the power amplifier or LNA, where loss is important (see Figure 5.3).
Table 5.3
Antenna Dissipative Losses
Location Component Typical loss (dB
Feed system Feed horn for reflector or lens 0.1
Waveguide series feed 0.7
Waveguide parallel feed 0.4
Stripline series feed 1.0
Stripline parallel feed 0.6
Active module at each element 0.0
Phase shifter Nonreciprocal ferrite, or Faraday rotator 0.7
Reciprocal ferrite 1.0
Diode (3- or 4-bit) 1.5
Diode (5- or 6-bit) 2.0
Diode (per bit) 0.4
Active module at each element 0.0
Array Mismatch (no electronic scan) 0.2
Mismatch (electronic scan 60º) 1.7
Exterior Radome 0.5-1.0
Source: [9, 44].
Fig ure 5.4 Series and parallel feeds. (After: [22].)
Mismatch loss also applies to phased arrays and is a loss due to impedance mismatch between the radiating elements of the antenna and free space. Mismatch loss is given by
where
Γ is the reflection coefficient and VSWR is the voltage to standing wave ratio [19, 20]. For a scanning array, the mismatch loss is given by [9]
where the element power gain is represented by
and 1 < β < 2 (usually ≈ 1.5) [9]. Given a VSWR of 1.5, mismatch loss for β = 1.25, 1.5, 1.75 and 2 is plotted in Figure 5.5. The average mismatch loss is 0.41, 0.66, 0.91, and 1.2 dB for β = 1.25, 1.5, 1.75, and 2, respectively (over 60º scan).
Fig ure 5.5 Mismatch loss vs. angle. (After: [9].)
As a note, some antenna analysts subtract the antenna losses from the antenna directivity (see Chapter 2) and term the result the antenna gain. Because of this, one must take care when using antenna directivity, antenna gain, and antenna losses in the radar range equation.
When the beam of a phased array antenna is scanned off of broadside (off of array normal), the antenna directivity decreases. If this is not explicitly included when generating the antenna pattern at the scanned angle, it should be included as a loss. Barton suggests a factor of
where θ is the scan angle [9, p. 369].2 Figure 5.6 contains a plot of (5.1) for β = 1.0, 1.5, and 2.0. The average scan loss is 1.3, 1.6, 1.9, and 2.2 dB for β = 1.25, 1.5, 1.75, and 2, respectively (over 60-deg scan).
The next loss we discuss is beamshape loss.3 This loss is associated with the situation where the antenna beam is not pointed directly at the target or where the beam is scanning across the target during the time the radar is coherently or noncoherently integrating a sequence of pulses (coherent and noncoherent integration is discussed in Chapter 8). In both cases, the full effect of the antenna directivity (GT and GR) terms of the radar range equation will not be realized. This most often happens during search. It is not applicable during track because it is assumed the target is very close to beam center during track.
Fig ure 5.6 Scan loss vs. angle. (After: [9].)
We account for both of the above situations by including beamshape loss as one of the loss factors. Historically, radar analysts have used the values of 1.6 or 3.2 that were derived by Blake in his 1953 paper [25]. However, Hall and Barton [5, 9, 26] revisited this problem in the 1960s and derived revised loss numbers of 1.24 and 2.48 dB. It should be noted that Barton and Hall indicate that there are many factors that affect scan loss, such as beam step size in phased array radars, number of pulses noncoherently integrated, whether or not the radar is continuously scanning, and detection probability. As such, the values of 1.24 and 2.48 dB should be considered rule-of-thumb numbers that would be suitable for preliminary radar analysis or design. In a more detailed analysis, these numbers should be revised based on the factors discussed by Barton and Hall.
The value of 1.24 dB is related to what is termed 1-D scanning, and the value of 2.48 dB relates to 2-D scanning. 1-D scanning would be associated with search radars that use a fan beam (a beam with a large beamwidth in one dimension (usually elevation) and a narrow beam in the other dimension). The radar would then rotate (or nod) in the narrow beam dimension but remain fixed in the wide beam direction. An example of such a radar is considered in Example 2 of Chapter 6, where we analyze a search radar with a cosecant squared elevation beam. In these fan beam types of radars, we assume the antenna directivity does not change much in the wide direction and that there is no need to include another loss. If this is not the case, we would want to use the 2-D beamshape loss.
In these situations, the antenna directivity (in the direction of the target) changes as the beam scans by the target, thus not all of the pulses will exhibit the same SNR. This, in turn, could affect the computation of detection probability (see Chapters 6 and 8). To account for this, we include the 1-D beamshape loss in the loss term of the radar range equation.
An example of where the use of the 2-D beamshape loss would be appropriate is in phased arrays radars (such as the second and third examples of Table 5.2) that scan a sector by stepping the beam in orthogonal directions (azimuth and elevation or u and v—see Chapter 12). In this situation, the radar would move to a beam position and transmit a pulse, or burst of pulses, and then move to another beam position. Because of this action, it is likely that the target could be off of beam center in two dimensions, thus the need for the 2-D beamshape
loss. In this situation, it may also be appropriate to include the scanning loss of (5.6) if the angular extent of the search sector is large.
A situation where we might want to use only a 1-D beamshape loss with a phased array radar is where we are generating a detection contour (see Example 2 of Chapter 6). In such a case, we would use the antenna directivity plot in, for example, elevation, and have the directivity as a function of elevation. However, we would need to account for the fact that the target is not on beam center in azimuth. Thus, we would include a 1-D beamshape loss in the radar range equation.
In discussing the phased array examples, we made the tacit assumption that the beams of the search sector were spaced close together as illustrated by Figure 5.7. This is similar to what Barton terms dense packing [9] and is characterized by the fact that there is no angular region that is not covered by the 3-dB beam contour of the radar (the 3-dB beam contours are the circles in Figure 5.7). Barton discusses another type of packing he terms sparse packing, wherein there may be parts of the angle space that are not covered by beams on any one scan (but hopefully will be covered on successive scans). In this situation, he points out that the beamshape loss now becomes a function of detection probability. This is something that should be considered in detailed studies of the impact of search methodology on detection performance of the radar.
We continue our previous example by adding antenna losses to Table 5.2 to generate Table 5.4.
We assumed the L-band search radar is a scanning radar with a cosecant squared beam.
Therefore, we included only the feed and 1-D scan loss. We assume that the S- and X-band radars are conducting a wide sector search and include mismatch (VSWR = 1.5) and scan losses we computed from (5.4) and (5.6), respectively, using β = 1.5 and θ = 30º, which is one-half the assumed ±60° extent of the search sector. We assumed the beams in the S-band radar were tightly packed and used Barton’s 2-D scan loss of 2.48 dB. For the X-band radar, we assumed the beams were not as tightly packed and thus used the historical 2-D scan loss of 3.2 dB. We assumed the radome on the S-band radar was cloth and use a fairly low value of radome loss. We assumed a hard radome on the X-band array and used a larger value of radome loss.
Fig ure 5.7 Examples of dense and sparse beam packing.
Table 5.4
Example Transmit RF and Antenna Losses