Varying modulation schemes
Recalling once again, SDR systems must inherently accommodate a variety of modulation schemes that vary in their demands and complexity within the same system architecture. Higher-order schemes, such as 64QAM, are employed when signal conditions permit in order to deliver higher data rates. To show the effect of using such modulation techniques, the 16QAM source in the previous example is replaced with a 64QAM source of the same bandwidth and pulse shaping. The software automatically configures the receiver to detect the 64QAM signal. Figure 4 shows the resulting BER simulation with phase noise using 64QAM with the same receiver RF characteristics described in Figure 6. The best achievable BER using 64QAM is 10-3.
BER evaluation of the receiver when presented with a 64QAM signal including the effect of phase noise.
64QAM receiver performance excluding the effect of phase noise.
The CCDF of both 16QAM and 64QAM signals.
The 64QAM BER results are shown in Figure 5 (with phase noise removed from the equation), which shows that a BER of 10-6 cannot be achieved. By viewing the complementary cumulative distribution function (CCDF) of 16QAM versus 64QAM in Figure 6, the peak-to-average ratio (PAR) of 64QAM exceeds that of 16QAM by about 1dB.
To safely compensate for the difference in PAR, the amplifier’s P1dB output is set to 14dB, and the resulting BER simulation (without the effect of phase noise) is shown in Figure 7. Now it is necessary to reintroduce phase noise into the system and to determine how much phase noise can be tolerated while still maintaining a BER of 10-6.
BER results using a signal with a P1dB RF power of 14dB.
The phase noise data file in VSS software consists of two columns of data. The first column represents offsets in frequency from the carrier (from 10Hz to 1MHz) and the second provides signal levels below that of the carrier. Each frequency point within the range of offsets is 10 times that of the preceding value. At a frequency offset of 10Hz, the phase noise mask is -31.6dBc/Hz, and at a 1MHz offset is -141.8dBc/Hz. To accommodate 64QAM, the phase noise values of the data file were decreased by 9dB. The results in Figure 8, using modified phase noise data, show that with an SNR of 28.5dB a 10-6 BER can be achieved. Figure 9 shows the received IQ plot at an SNR of 30dB, which identifies a small amount of phase-noise-induced jitter that produces errors. The 16QAM BER simulation results using a modified phase-noise mask indicated that 10-6 BER can be achieved with an SNR of about 21dB.
BER results from a signal with P1dB power of 14dB and modified phase noise data.
Figure 9. The IQ plot of 64QAM with an SNR of 30dB.
Importance of versatility
Simulation software such as VSS can be used to evaluate complex SDR receivers that must process the simplest FM waveforms as well as higher-order modulation schemes of 64QAM. Fortunately, in addition to the example above, it is possible to identify other potential signal-degrading factors, such as IQ imbalance, and the impact of interfering signals on BER. Complex communications systems require analysis at all levels of the design process in order to achieve desired performance. While the ultimate goal of SDR--to implement the entire radio design in software and minimize the amount of the hardware needed--is still in the future, increased capabilities in simulation solutions is helping the growth of SDR implementations in both military and commercial networks in important ways.
If you liked this article...
About the Author:
Joel Kirshman is an applications specialist in AWR’s Visual System Simulator software and has been with the company for more than a decade. Prior to joining AWR, Joel worked for Elanix in a similar role as a system simulation expert.
- Head to the RF and Microwave Designline homepage for the latest updates in RF and microwave.
- Sign up for the weekly RF and Microwave Designline Newsletter to be delivered to your mailbox with the latest highlights from the site.
Joel holds a master degree in electrical engineering from California State University, Northridge. Prior to entering the EDA industry, Joel spent a number of years at The Aerospace Corporation where he worked on various project from designing a digital processing unit for a space application, to rad-hard testing of electronic devices, and working for the GPS program office