This presentation by Muhammad Ismile from Loughborough University investigates the feasibility of using a low-cost Software-Defined Radio (SDR), the USRP B205 mini, for high-precision pulse characterization, a task typically requiring expensive equipment. The research methodology involved transmitting a pulse from the USRP at a fixed 10 MSa/s rate and capturing it with a high-speed oscilloscope at varying sampling rates, from 100 MSa/s up to 10 GSa/s.
The study demonstrates that significant oversampling at the receiver dramatically improves key pulse metrics. The measured rise time improved from 790 nanoseconds to 111 nanoseconds, approaching the theoretical ideal of 100 ns. Similarly, timing jitter was substantially reduced, becoming nearly negligible at sampling rates of 1 GSa/s and above. However, these ultra-high sampling rates introduce more system noise. The research concludes that a receiver-to-transmitter sampling ratio of at least 10:1, combined with post-processing techniques like averaging and filtering, is optimal for balancing signal fidelity and noise. This confirms that affordable SDRs can be viable for high-accuracy applications like radar and pulse communication when paired with appropriate high-bandwidth measurement and analysis techniques.
Achieving High-Precision Pulse Characterization with a Low-Cost SDR
00:00:59 Research Goal: The study aims to determine if a low-cost Software-Defined Radio (SDR), the USRP B205 mini, can achieve sub-nanosecond precision for pulse characterization, which is crucial for applications like radar and advanced wireless communications.
00:05:39 Experimental Setup: A pulse is generated in GNU Radio, transmitted by the USRP B205 mini at 2.4 GHz, and captured by a high-speed Lecroy oscilloscope capable of sampling at up to 18 Giga-samples per second (GSa/s).
00:07:30 Test Parameters: The experiment focused on a 10-microsecond pulse transmitted at a fixed rate of 10 Mega-samples per second (MSa/s). The receiver's sampling rate was varied from 100 MSa/s up to 10 GSa/s to observe its effect on signal fidelity.
00:09:06 Rise Time Improvement: A key finding was that increasing the receiver sampling rate dramatically improved the pulse's rise time. It decreased from 790 ns at 100 MSa/s to 111 ns at 10 GSa/s, approaching the theoretical ideal of 100 ns.
00:12:28 Jitter Reduction: Timing jitter, or the variation in pulse timing, was also significantly reduced by oversampling. At 1 GSa/s and 10 GSa/s, jitter became "nearly negligible," confirming highly consistent pulse alignment.
00:15:43 The Role of Filtering: At high sampling rates (β₯1 GSa/s), applying digital filters was effective in reducing system noise without distorting the pulse shape. However, at lower sampling rates, filtering caused significant signal distortion.
00:19:32 The Noise Trade-off: While ultra-high sampling rates improve temporal resolution, they also capture more system noise, which can degrade signal clarity in the steady-state portion of the pulse.
00:20:24 Optimal Sampling Ratio: The research concludes that a receiver-to-transmitter sampling ratio of at least 10:1 provides the best balance between capturing high-fidelity pulse edges and managing system noise.
00:20:57 Key Takeaway: Low-cost SDRs like the USRP B205 mini are viable for high-precision applications, but this requires using a receiver with a significantly higher bandwidth (oversampling) and employing post-processing techniques like averaging and filtering to achieve optimal results.