Receiver Bandwidth
Figure 1: Variation in signal amplitude within a bandpass filter with a −3 dB bandwidth
Figure 1: Variation in signal amplitude within a bandpass filter with a −3 dB bandwidth
Figure 1: Variation in signal amplitude within a bandpass filter with a −3 dB bandwidth
Receiver Bandwidth
Bandwidth
The bandwidth B, BW, or Δf is the difference between the upper and lower cutoff frequencies of a radar receiver and is expressed in hertz (Hz). For a baseband signal whose spectrum begins at zero frequency, the bandwidth is equal to the upper cutoff frequency. In a radar receiver, the bandwidth is usually determined by the bandpass filters of the intermediate-frequency amplifier. The receiver must be capable of processing the entire bandwidth of the received echo signal.
However, the greater the bandwidth, the greater the amount of background noise that also passes through the receiver. Since this noise is statistically distributed across the frequency spectrum, increasing the bandwidth also increases the received noise power. As a result, the signal-to-noise ratio deteriorates, reducing the achievable receiver sensitivity.
The frequency spectrum of a rectangular pulse can be analyzed using the Fast Fourier Transform (FFT). An ideal rectangular pulse theoretically contains frequency components extending to infinitely high frequencies. In practice, however, the characteristics of the transmitter, receiver, and available measuring instruments limit the usable frequency range. The greater the receiver bandwidth, the more accurately the original pulse shape with its steep edges can be reproduced.
For simple, unmodulated radar pulses, the required bandwidth B can be approximated by the reciprocal of the pulse duration τ. The exact value depends on the pulse shape and the definition of bandwidth being used:
| B ≈ | 1 | (1) |
| τ |
For moving targets, the Doppler effect causes a shift in the frequency spectrum of the reflected signal. For targets approaching the radar, the pulse duration is also shortened slightly, and the signal bandwidth increases accordingly. However, when determining the receiver bandwidth, it is particularly important to consider that the Doppler frequency shift moves the received signal spectrum relative to the receiver's passband response. Therefore, the receiver bandwidth must accommodate not only the signal bandwidth itself but also the expected Doppler frequency shifts.
Practical Example: Receiver Bandwidth of a P-12 Radar
During my service as commander of a P-12 radar station, I observed that reducing the bandwidth of the intermediate-frequency amplifier noticeably reduced the background noise. At the same time, however, the target indications of particularly fast aircraft, such as the SR-71, became weaker. Thus, a reduction in noise power did not necessarily improve the detectability of all targets.
One possible explanation is the Doppler frequency shift of the echo signals. For very fast-moving targets, part of the shifted signal spectrum may fall outside the passband of a narrowband intermediate-frequency filter. Consequently, not only the noise but also the desired signal is attenuated. The optimum receiver bandwidth is therefore always a compromise between noise suppression and receiving the desired signal with as little loss as possible.
In radar systems employing intrapulse modulation of the transmitted pulse, the receiver bandwidth is much greater than the reciprocal of the pulse duration. In this case, the required receiver bandwidth depends on the type of intrapulse modulation, the weighting function, and the duration of the compressed pulse. With linear frequency modulation, a greater bandwidth allows smaller range cells and improved range resolution, while avoiding excessively high sidelobes (time sidelobes) during pulse compression.
Scientific radars with particularly demanding range-resolution requirements use signal bandwidths on the order of 100 to 200 MHz in practice. Some sophisticated receivers also provide continuously adjustable bandwidth controlled by the radar signal processor.
Time-Bandwidth Product
A commonly used quantity for characterizing pulse compression is the time-bandwidth product τ·B (expressed in µs·MHz, yielding a dimensionless value), which is the product of pulse duration and signal bandwidth. In different radar systems, the time-bandwidth product may have values between 5 and 1000. For relatively small values, such as 5 to 15, sidelobe suppression of up to 35 dB can be achieved under favorable conditions. For larger values between 15 and 500, carefully optimized systems can achieve sidelobe suppression of up to 45 dB. However, these values should not be regarded as generally achievable performance levels. The actual sidelobe suppression depends significantly on the modulation method, weighting function, and overall signal processing, and is substantially lower in many practical applications.
See also: examples of modulation methods for pulse compression.
With sufficiently large signal bandwidth, pulse compression can produce pulses only a few nanoseconds in duration. This makes range resolution of better than one meter possible. If the propagation times of the direct signal and the signal reflected from the Earth's surface differ sufficiently, separate processing of the two signals can also be used for height determination.