Bit Error Rate (BER) Calculator
Calculate bit error rate from Eb/N0 SNR for BPSK, QPSK, 16-QAM, 64-QAM, and 256-QAM modulation. AWGN-channel theoretical performance.
How to Use
- Enter Eb/N0, the energy per bit over the noise density, in dB.
- Choose the modulation: BPSK, QPSK, 16-QAM, 64-QAM or 256-QAM.
- Read the bit error rate, the expected errors per gigabit and how far you are above the Shannon limit.
- Compare modulations on the chart: each curve is the uncoded error rate on a clean noise-only (AWGN) channel.
Worked Example
A BPSK link at an Eb/N0 of 10 dB. 10 dB is a ratio of 10, so BER = Q(√(2 × 10)) = Q(4.47) = 3.9 × 10⁻⁶: about 3,900 errors in every billion bits, before any error correction.
Switching to 16-QAM at the same 10 dB carries four times the bits per symbol but raises the BER to about 1.8 × 10⁻³; getting back to 10⁻⁶ needs 14.4 dB, roughly 4 dB more power.
The common mistake: entering the SNR as Eb/N0. They differ by the ratio of bandwidth to bit rate. For 16-QAM at one symbol per hertz, the bit rate is four times the bandwidth, so Eb/N0 is 6 dB below the SNR, and the real BER is far worse than it looks.
Show Work
Formulas
History of BER Analysis
Claude Shannon's 1948 paper established the theoretical minimum Eb/N0 for reliable communication at any bit rate - the Shannon limit. Uncoded modulation needs far more: BPSK needs 9.6 dB for a BER of 10⁻⁵, about 11 dB above Shannon's limit. Modern error-correcting codes (Viterbi 1967, Turbo 1993, LDPC 1963/1996, Polar 2008) have closed the gap to within 1 dB of Shannon, enabling satellite links, 5G cellular, and DVB broadcasting at near-theoretical limits.
About This Calculator
Enter Eb/N0 (bit-energy-to-noise-power-spectral-density) in dB and pick modulation. The tool computes theoretical BER for AWGN channel using closed-form BPSK/QPSK formulas and M-QAM approximations.
Real-world BER is worse due to multipath, interference, phase noise, and sync error. Modern coded systems (LDPC, Turbo, Polar) can operate within 1 dB of Shannon limit, dramatically reducing required Eb/N0. Everything runs client-side.
Reading the curve and choosing a modulation
The most important habit when reading a BER curve is to notice how steep it is. Because the bit-error rate follows the tail of a Gaussian, a small change in Eb/N0 produces an enormous change in errors: near a usable operating point, adding just 1 dB of signal-to-noise can drop the error rate by an order of magnitude or more. That steepness is why link budgets are fought over in fractions of a decibel, and why a link that looks marginal at a BER of 10⁻³ can become rock-solid at 10⁻⁶ with only a couple of extra dB of margin from a better antenna, a lower-noise front end, or a shorter range.
Picking a modulation is a trade between spectral efficiency and required power. Denser constellations such as 16-QAM, 64-QAM and 256-QAM pack more bits into each symbol, so they carry more data in the same bandwidth — but the constellation points sit closer together, so each step up demands several more dB of Eb/N0 to hold the same BER. BPSK and QPSK are the rugged choices for weak-signal and long-range links such as deep-space, satellite uplinks, and control channels, while high-order QAM rules short, clean, high-throughput links like cable, close-range Wi-Fi, and the strong-signal resource blocks of a 5G connection. Adaptive systems simply switch between them on the fly, dropping to a lower order as the channel degrades so the link stays up instead of dropping out.
Frequently Asked Questions
What is the difference between Eb/N0 and SNR?
SNR compares total signal and noise power in the receiver bandwidth; Eb/N0 compares the energy of one bit with the noise in 1 Hz. They are related by Eb/N0 = SNR × B/R. For QPSK sent at one symbol per hertz, the bit rate is twice the bandwidth, so Eb/N0 is 3 dB below the SNR.
What BER do I need?
Uncoded voice links tolerate about 10⁻³. Data links usually aim for 10⁻⁹ to 10⁻¹² after error correction, and modern systems deliberately run the raw link at 10⁻³ to 10⁻² and let a strong code clean it up.
What is the Shannon limit?
The lowest Eb/N0 at which error-free data is possible at all: −1.59 dB, reached only with unlimited bandwidth. At 1 bit/s per hertz the limit is 0 dB. Uncoded BPSK needs 9.6 dB for 10⁻⁵, about 11 dB more, and good codes close most of that gap.
Why does QPSK have the same BER as BPSK?
QPSK is two BPSK signals on carriers 90° apart. Each bit gets the same energy and the same noise as in BPSK, so the bit error rate is identical, while twice as many bits fit in the same bandwidth.
How much more power does each QAM step need?
For a BER of 10⁻⁶: QPSK 10.5 dB, 16-QAM 14.4 dB, 64-QAM 18.8 dB, 256-QAM 23.5 dB. Each step adds 2 bits per symbol and costs roughly 4 to 5 dB.
How do I use the Bit Error Rate (BER) Calculator?
Just type your numbers. The answer shows up right away — there is no button to press. Change anything and it updates by itself.
Is it free? Does it work without internet?
Yes to both. It is free with no sign-up, and once the page has loaded it keeps working even with no internet.
Where does my data go?
Nowhere — every calculation runs on your own device. Nothing you enter is uploaded, logged, or stored.
Common Use Cases
Wi-Fi and Cellular
At 15 dB Eb/N0, uncoded 64-QAM has a BER of about 7.7 × 10⁻⁴; the error-correcting code turns that into clean data, and the link drops to 16-QAM or QPSK when the signal weakens.
Satellite Links
Uncoded QPSK needs 9.6 dB for 10⁻⁵; at 8 dB it is 1.9 × 10⁻⁴, so standards such as DVB-S2 rely on strong LDPC codes to work well below that.
Deep-Space Links
At 2 dB, uncoded BPSK makes 3.8 errors per hundred bits. Turbo and LDPC codes are what make links this weak usable.
Cable and Fibre Modems
256-QAM needs about 23.5 dB for 10⁻⁶ uncoded, which is why it is used on clean wired links rather than over the air at long range.
Last updated: