SNR
Understanding SNR
SNR is the master metric of RF system performance. Every other performance metric, including BER, EVM, channel capacity, and detection probability, is ultimately determined by SNR. Maximizing SNR is the primary goal of receiver and system design.
SNR and Shannon Capacity
Shannon's theorem states the maximum achievable data rate through a noisy channel: C = BW x log2(1 + SNR). This theoretical limit shows that both bandwidth and SNR determine channel capacity, and that there is a fundamental tradeoff between them.
Improving SNR
- Increase signal power: Higher transmit power or antenna gain increases received signal.
- Reduce noise: Lower noise figure receiver, narrower bandwidth filter.
- Processing gain: Spread spectrum, coherent integration, coding gain.
- Interference mitigation: Spatial filtering (beamforming), frequency hopping.
SNR (linear) = P_signal / P_noise
Shannon capacity:
C = BW x log2(1 + SNR) bits/second
Required SNR for BER = 10^-6:
BPSK: 10.5 dB
16-QAM: 14.5 dB
64-QAM: 18.5 dB
256-QAM: 24.5 dB
Frequently Asked Questions
What is SNR?
SNR is the ratio of signal power to noise power (in dB). Higher SNR means better signal quality. SNR determines the maximum data rate, achievable modulation order, and bit error rate. It is the most fundamental metric in RF system performance.
What SNR is needed for good communication?
It depends on the modulation scheme. BPSK needs about 10 dB for BER = 10^-6. 64-QAM needs about 18 dB. 256-QAM needs about 24 dB. Voice communications over FM can work with as little as 12 dB SNR, while 4K video streaming over Wi-Fi needs 30+ dB.
How do you measure SNR?
SNR can be measured with a spectrum analyzer by comparing the signal power to the noise floor. For modulated signals, EVM measurement provides an equivalent SNR estimate. Purpose-built signal analyzers can separate signal from noise using reference signals or training sequences.