What Is Signal to Noise Ratio?

Signal to noise ratio (SNR), also written S/N, is a dimensionless measure that quantifies the relative strength of a desired signal compared to the background noise present in the same channel, circuit, or measurement. A high SNR indicates that the signal is clearly distinguishable from noise; a low SNR indicates that the signal is difficult to detect or accurately recover. The metric is ubiquitous in electrical engineering, communications, audio, radar, and scientific instrumentation, serving as a fundamental figure of merit for any system that must extract information from a noisy environment.

SNR is defined as the ratio of signal power to noise power. Because the range of practical values spans many orders of magnitude, engineers most often express SNR on a logarithmic scale in decibels (dB): SNR (dB) = 10 × log₁₀(P_signal / P_noise). A ratio of 10:1 corresponds to 10 dB; 100:1 corresponds to 20 dB; 0 dB means signal and noise power are equal. The IEEE Transactions on Signal Processing has published foundational and applied research on SNR estimation, measurement, and the design of systems that must operate at low SNR margins.

Definition, Measurement, and Noise Figure

The definition of SNR requires careful specification of what constitutes the signal and what constitutes the noise, since both concepts depend on the application context. In a communications receiver, the signal is the transmitted baseband waveform after demodulation, and the noise includes thermal noise, interference, and quantization error. Noise figure (NF) is a closely related parameter that characterizes how much an amplifier or receiver chain degrades the SNR of the input signal; a noise figure of 3 dB means the component halves the SNR. Discrete-time measurement of signal-to-noise power ratio poses specific sampling and averaging requirements, addressed in IEEE research on discrete-time SNR measurement methods that examines how sampling rate, window length, and spectral leakage affect measurement accuracy. The noise floor of a system sets the minimum detectable signal level: signals weaker than the noise floor cannot be reliably recovered regardless of subsequent processing.

Improving SNR through Filtering

Filtering is one of the primary techniques for improving effective SNR by selectively attenuating noise while preserving the signal. A matched filter, the theoretically optimal linear filter for detecting a known waveform in additive white Gaussian noise, maximizes the output SNR at the sampling instant. Its transfer function is the complex conjugate of the signal's spectrum in the frequency domain. Bandpass and narrowband filters improve SNR by excluding noise energy outside the signal's frequency band, which is effective whenever the signal and noise occupy different spectral regions. Adaptive filters such as the least mean squares (LMS) algorithm continuously adjust their coefficients to track changes in the noise environment, making them useful in echo cancellation, active noise control, and interference rejection. The NIST technical notes on signal measurement and calibration describe how filtered measurements are validated against reference standards in precision metrology applications.

Diversity and Multi-path Reception

In wireless communications, the received SNR fluctuates as the channel changes due to multipath propagation, shadowing, and Doppler shifts. RAKE receivers, used in code-division multiple access (CDMA) systems, improve effective SNR by resolving individual multipath components and coherently combining them, exploiting the energy that would otherwise be lost to inter-path interference. Each finger of the RAKE correlates the received signal against a time-shifted copy of the spreading code, aligning the multipath echoes so that their amplitudes add constructively. Antenna diversity, in which multiple physically separated receive antennas capture independent faded copies of the signal and the strongest or a weighted combination is selected, provides a complementary approach to maintaining adequate SNR in mobile environments.

Applications

Signal to noise ratio has applications across a wide range of disciplines, including:

• Wireless communications, where link budget analysis determines the transmit power and antenna gain required to meet a target SNR at the receiver
• Audio equipment design, where SNR specifies how cleanly a microphone, amplifier, or digital recorder reproduces sound
• Radar and sonar, where the probability of target detection is directly governed by the received SNR after pulse compression
• Medical imaging, including MRI and ultrasound, where SNR determines image resolution and diagnostic quality
• Scientific instrumentation, where spectroscopic and sensing measurements require sufficient SNR to resolve small signal variations