Download this article in .PDF format
This file type includes high resolution graphics and schematics when applicable.

Wireless technology is so widespread that the frequency spectrum can be quite crowded. Almost every portion of available frequency spectrum is shared by multiple services and/or applications, requiring practical, effective control of interference. To eliminate interference, it must first be isolated and identified. Hunting down sources of interference can be challenging, if not nearly impossible.

One example is trying to find short-duration intermodulation signal products without the right tools, or without  knowledge about antenna types, signal characteristics, and other factors. Hopefully, this overview of interference-locating techniques and some examples will help accelerate the success of future interference-hunting efforts.

Preselector energy

Locating interference starts with finding the right measurement tool, a process which can be aided by understanding the differences between different spectrum analyzer types. Two common spectrum analyzer types are swept-tuned spectrum analyzers and real-time spectrum analyzers. Swept-tuned spectrum analyzers are typically based on a superheterodyne receiver architecture in which signals are mixed with a tunable local oscillator (LO) to cover a bandwidth of interest from a starting frequency to a stop frequency.

A real-time spectrum analyzer samples a portion of the frequency spectrum and uses digital-signal-processing (DSP) techniques to analyze the captured spectrum. Filtering occurs in both approaches, and filtering helps set an analyzer’s resolution bandwidth (RBW), by which signals closely spaced in frequency can be isolated and identified.

Interference can cause degraded system performance. For example, the unwanted energy may be causing coverage, reception, or access problems for a communications system. Problems can include an adequate signal but poor reception; an adequate signal but intermittent or no access; or a poor signal and no reception. Three basic types of interference are the most common:

  • Co-channel interference is one of the simplest forms of interference, where more than one transmitter can be found on the same channel. This is not unusual, given that many frequencies and frequency bands are at least partially shared by numerous applications or users.
  • Adjacent-channel interference results from energy originating from a transmitter other than the one intended. Such a secondary transmitter could be geographically close or close in operating frequency, and operate at a much higher power level than the intended transmitter.
  • Intermodulation-based interference occurs when energy from two or more transmitters mixes together to create spurious signal products or frequencies. In general, third-order spurious products are the most common intermodulation interference caused by only two sources, although this type of interference can be caused by more than two sources. Transmitters that produce this type of interference are usually close together and at higher power levels.

Figure 1 shows the simplified function of a preselector filter within a receiver. Although the filter is tuned to remove interference, interference can occur in a receiver when there is energy inside the preselector’s filter bandwidth. Suitable energy within the preselector bandwidth can impact receiver performance; it does so by blocking the receiver directly from detecting a desired signal, or else through a form of desensitization in which lower-level signals are not detected.

The logarithm scale

Protocol-based test tools have often been used to identify interference. While they are useful in finding interference, such tools are also limited in providing insight into how a system might be impacted by interference and in identifying the types of interference that can cause the most harm to a system. For such purposes, a spectrum analyzer provides a clear view of a given portion of spectrum, thus pinpointing where interference might fall relative to a communications system of interest.

When seeking a spectrum analyzer for finding interference, a key specification to consider is probability of intercept. This is simply the minimum duration of an interfering signal and the chance that the analyzer can detect and display it. Swept-tuned spectrum analyzers typically have a low probability of intercept, meaning that the interfering signal would have to be present for tens of milliseconds. Real-time spectrum analyzers, have a high probability of intercept to their maximum span, with the capabilities of detecting signals as brief as microseconds. Such analyzers can identify signals that are of much shorter duration than a traditional swept-tuned spectrum analyzer.

RBW settings

Spectrum analyzers show captured signals on a logarithmic (rather than linear) scale, with signals and interference displayed in decibels (dB) or decibels relative to 1 mW (dBm) power. The logarithmic scale makes it possible to show a much wider dynamic range on an analyzers screen than a linear scale.

There are a few tricks to relate dBm values to a linear difference (Fig. 2). Because it is logarithmic, every 10-dB change in power is a power of 10 change in wattage. Similarly, a 3-dB change in power represents a doubling (up 3 dB) or halving (down 3 dB) of wattage. Measurement of +30 dBm is the same as measuring 1 W, while measuring +33 dBm is the same as measuring 2 W. This can be significant, since most spectrum analyzers are limited to about +30 dBm input power, and feeding an analyzer excess power can damage the instrument.

For interference applications, a spectrum analyzer’s RBW control is very important. It is a filter that helps discriminate between wide- and narrow-bandwidth signals in the same span by changing the RBW value (Fig. 3). If the RBW is set too wide [Fig. 3 (left)], smaller signals close to larger signals will be lost. A narrow RBW filter can easily discriminate between two signals that are close together, but it will slow down the spectrum analyzer. This requires a longer measurement duration to ensure signal probability of intercept.

Download this article in .PDF format
This file type includes high resolution graphics and schematics when applicable.