By Robert M. Lund, Comcast; Kathryn Sanders, Sanders RF Consulting LLC
Recent advances in digital communications algorithms and artificial intelligence, coupled with the exponential growth of personal computing power, have positioned software-defined radio (SDR) technology as a viable tool for RF signal capture and playback, simulation, and testing.
Popular applications for modern SDR designs include:
- Cellular handsets with programmable digital signal processing (DSP)
- University research/teaching aids
- Military and satellite communications platforms that make use of programmable cores for intermediate frequency (IF) and baseband signal processing
- Cognitive radio – a “smart” radio mesh that can adapt to interference adjusting frequency or power for example
- Amateur radio
The goal of this paper is to outline the general architecture and capabilities of a software-defined radio, followed by some basic principles of digital signal processing. A familiarity with sampling theory and the discrete Fourier transform, for example, will be very helpful for anyone looking to get started with SDR.
Finally, two lab applications of software-defined radio will be detailed. Both of these applications were used in testing a profile management application implementation, so a section on PMA is introduced. After familiarization with PMA, the first test example uses SDR to capture long term evolution (LTE) signals, then re-play them back into the RF plant to be subsequently detected by pattern recognition software that is a function of PMA. The second case recreates plant conditions from a cable modem’s perspective by translating its reported receive modulation error ratio (RxMER) values into a waveform to be used as an impairment profile applied to an orthogonal frequency division multiplexing (OFDM) channel. Again, this impairment can then be analyzed by the analytics engine of PMA, and a custom OFDM profile recommendation can be verified.