Communication ES Signal Processing

Communication ES Signal Processing. The classic communication ES receiver implementation is basically a high-quality manually controlled superheterodyne receiver. Signal search was performed by the operator manually tuning the receiver through the frequency range known to be used by the adversary’s radios and listening to the outputs of the available demodulator(s) for signals of interest.

When such a signal was found, the operator would listen to the demodulated signal and record the observations. If available, a DF system would be tuned to the frequency and measurements obtained for the signal angle of arrival. This process required the attention of a skilled operator and had the additional weakness that short duration transmissions on new frequencies could be missed, particularly if the frequency ranges to be covered could not be divided up among multiple systems and operators. Another weakness concerned the size, weight, and power consumption of the equipment.

Modern purpose-designed communication EW receivers provide significant enhancements:
1. Computer controlled operation via standard digital interfaces;
2. Accurate high-speed tuning and reduced phase noise that results from the use of high-quality crystal oscillators as frequency references and sophisticated frequency synthesis techniques;

3. Provisions for phase coherent operation of multiple receivers to allow commonality of hardware between systems used for signal search and DF;
4. Built-in-test functionality;
5. Reduced size, weight, and power consumption.

Digital signal processing techniques are being adopted for advanced ES systems. Digital filter bank concepts based on the Fast Fourier Transform algorithm allow a single wideband receiver to process and detect the individual signals present within a large instantaneous bandwidth. Also, if the system dwells on a fixed center frequency, digital down-converters can be used to extract the narrowband signals within the receiver bandwidth and software demodulators used to recover the message content from each signal.

Advanced wideband communication ES sensors based on digital filter bank techniques have some very desirable advantages:

1. A large frequency range can be scanned quickly; the tuning frequency step size can be orders of magnitude larger than the required frequency resolution. This method substantially reduces or eliminates the likelihood that a short duration transmission will be missed and can provide some capability for detecting at least some hops transmitted by a frequency hopping radio;

2. The use of Constant False Alarm Rate techniques allows the system detection processing parameters to be adjusted automatically to achieve the best possible sensitivity without incurring erroneous signal detections at a rate that exceeds a set value, even if the environmental noise is frequency dependent and time variant (18);

3. Algorithms can be implemented to determine the type of modulation used by a signal and the modulation parameters;
4. Raw signal data can be acquired and stored for off-line analysis;

5. Demodulators implemented in software can accommodate a wide range of modulation types;
6. DF functionality can be integrated into the system to provide a measurement of the angle of arrival for each signal that is detected;

7. Reports of signal detections and the measured signal parameters can be automatically stored in a database and transferred to EW analysis and intelligence systems for subsequent processing;
8. Remote controlled or autonomous operation of ES systems is feasible.

However, wideband signal processing techniques also incur disadvantages. Early implementations tended to be expensive and have significant performance limitations. A major problem concerns dynamic range, which is a measure of the ability of a system to process strong and weak signals simultaneously.

This issue is of considerable importance for wideband communications ES systems because weak signals of interest and strong signals will often coexist in the same frequency range. The dynamic range of a practical system is dependent on the noise and spurious signals, which are generated in the system by various mechanisms. One of the most important of these mechanisms is third order intermodulation distortion.

This occurs when two or more signals present within the system bandwidth interact because of nonlinearities in the system signal processing. The spurious signals that result remain within the system bandwidth and, depending on the size of the input signals and the nature of the system nonlinearities, can be large enough to be detected and interpreted as actual signals in subsequent processing.

To avoid this undesirable result, the detection processing must be adjusted to reduce the effective system sensitivity. Thus, the presence of strong input signals tends to degrade the ability of the system to detect and process weak signals usefully. The problem is aggravated as the system bandwidth is increased because the number of strong signals within the system bandwidth can also be expected to increase.

Fortunately, progressive advances in radio frequency components, analog-to-digital converters, and digital processor hardware have substantially resolved these issues, particularly when careful system design choices and tradeoffs are made. Nevertheless, a well-designed narrowband receiver may still offer advantages with respect to usable sensitivity and selectivity in a dense signal environment that includes strong signals.

In addition to its message content, a communication signal contains information that can be used to classify the type of signal, and, with some limitations, to identify individual emitters.

The measurement of the modulation type and parameters is an important topic for communications ES systems. Conventional communication systems use modulation techniques to embed information on a sinusoidal carrier signal. The choice of modulation type and implementation parameters is dependent on application requirements and various factors, such as the need for interoperability with other radio systems as well as technology and cost constraints.

Advances in communication theory coupled with the availability of low-cost digital signal processing hardware have motivated the use of sophisticated digital modulation techniques to provide favorable trade-offs between bandwidth efficiency, sensitivity to propagation effects, and hardware implementation costs. At the same time, simple, classic modulation techniques, such as analog frequency modulation, remain in widespread use, in part to maintain interoperability with older systems.

Knowledge of the modulation type and parameters associated with a signal is of considerable practical value. Requirements for interoperability have led to the standardization of the modulation types used by military radios. For example, the tactical VHF radios used in ground operations typically support analog FM and digital FSK modulations in accordance with standards such as MIL-STD-188- 242. If a signal has a modulation type and parameters associated with a communication system known to be used by an adversary, then it can be flagged as a potential signal of interest and prioritized to receive attention.

Also, because emitters that are communicating with each other will generally use the same modulation type, this knowledge can be used to support or reject hypotheses that concern the membership of a given emitter in a network. Finally, knowledge of the modulation type and parameters facilitates the selection of an appropriate demodulation technique to recover the message content.

Because of the diversity of modulation standards and the effects of multipath propagation and nonideal radio system implementations, the modulation recognition problem is nontrivial. Algorithms for modulation recognition have been described in various papers, of which Refs. 19-22 are representative examples.

A related idea is based on the observation that the signal waveforms generated by practical radio transmitters will differ in subtle ways depending on implementation details and component tolerances, and that these differences can be sufficient to distinguish between transmitters that are very similar or even nominally identical. Various techniques have also been proposed to extract and measure appropriately selected features from a signal and use statistical tests to determine whether the feature measurements match those of previously observed signals.