6 SUMMARY AND RECOMMENDATIONS
The key factors in radio receivers where digitization occurs at the IF or RF are analog-to-digital
conversion and digital signal processing. Although not discussed in this paper, digital-to-analog
conversion is also a factor for applications requiring analog output (such as voice). Because of
the rapid advances in hardware development of ADCs, DACs, digital signal processors, and
specialized ICs, development of radio receivers using digitization at the IF (and in some cases at
the RF) is becoming increasingly popular.
Hardware limitations of ADCs, digital signal processors, and DACs place constraints on
digitization at the IF and RF in radio receivers. It was shown that digitization at the RF, in
general, requires some type of bandlimiting (filtering) and amplification before the actual
digitization takes place. The required amount of filtering and amplification is application-
specific. ADC performance is improving rapidly. However, there is a tradeoff; one can get either
high sampling rates or high resolution, but not both simultaneously. Therefore, the high sampling
rate ADCs required for wide bandwidth applications may not have sufficient SFDR. Digitization
at the RF is now being considered for satellite receivers since a large SFDR is not a necessity,
very high sampling rate ADCs already exist, and even faster ADCs are being developed. For
receiver applications requiring a large SFDR, such as HF communications, digitization at the IF
is currently a more practical option.
Digital signal processors may present an even greater limitation than ADCs in radio receivers
using digitization at the RF or IF. The speed, size, and cost of these processors are important for
a particular radio receiver application. The requirement for real-time operation for many radio
receivers places a heavy burden on these processors. It is difficult to discuss limitations of digital
signal processing in general terms since many algorithms can be implemented in radio receivers
depending on the specific application. The amount of time that signal processing requires is a
function of the bandwidth of the signal, the speed of the processor, and the number and
complexity of the algorithms required to perform the needed radio receiver functions.
Several potential methods and devices that are expected to be useful in radio receivers employing
digitization of the RF or IF are discussed in this report. Various types of quantization techniques
include uniform, μ-law, adaptive, and differential quantization. These techniques can be used to
improve the dynamic range of ADCs as well as make the SNR insensitive to input signal
amplitude. Nonlinear compression devices such as log amplifiers and automatic gain control
amplifiers also can be used to improve the dynamic range of ADCs. Postdigitization algorithms
for improving the SFDR provide extended dynamic range for presently available ADCs.
Sampling downconverters, based on the theory of bandpass sampling, may be coupled with ADC
technology to provide improved ADC performance for bandpass sampling applications.
Specialized ICs for digital signal-processing tasks required in radio receiver applications show
great potential in increasing processing speed; this allows more receiver functions to be executed
in real time. Bandpass sampling is expected to be important for radio receiver applications since
it allows the use of more readily available ADCs with lower sampling rates and higher SFDR.
This report provides information on receivers implementing digitization at the RF or IF. The
topic of EMC analysis for these radio receivers needs to be investigated further. Due to the
tremendous variety of possible receiver implementations employing digitization at RF or IF,
44
focusing on some specific RF front-end configurations probably would be the best approach for
determining EMC analysis procedures. In general, however, RF front-end specifications such as
SFDR and intermodulation distortion should be important, just as in EMC analyses of traditional
receivers not employing digitization of the RF or IF.
The potential for spectrum overlap due to the sampling process is probably the major difference
between receivers that employ digitization at the RF or IF and those that do not. Effects of this
spectrum overlap are best determined by considering the specific radio system details such as the
type of source information (i.e., voice, data, or video); desired signal bandwidth; modulation and
coding techniques; undesired signal characteristics (bandwidth, power, and type of signal); and
the performance criterion used to evaluate the quality of the reception of the desired signal. The
sampling rate in relationship to the maximum frequency content of the signal and the signal
bandwidth is of utmost importance in evaluating spectrum overlap. Computer simulation is
probably the best tool available to evaluate the effects of spectrum overlap on receiver
performance.
45
Chia sẻ với bạn bè của bạn: |