
C/N, C/N0, and Eb/N0 Explained | Core Signal Quality Metrics in Satellite Communication
Engineering guide to C/N, C/N0, and Eb/N0 in satellite systems covering definitions, formulas, relationships, DVB-S2 examples, rain fade effects, and common mistakes in signal quality analysis.
Every satellite link ultimately comes down to one question: is the signal strong enough relative to the noise to deliver usable data? Three metrics answer that question at different levels of abstraction — C/N, C/N0, and Eb/N0 — and every satellite engineer needs to know when to use each one, how they relate, and where the conversions go wrong.
You will encounter these three metrics everywhere in SATCOM. C/N appears on modem lock indicators, spectrum analyzer displays, and link budget spreadsheets. C/N0 shows up when comparing links that use different bandwidths or data rates. Eb/N0 appears in modulation performance curves, FEC threshold tables, and DVB-S2 MODCOD selection logic. Together, they form the chain that connects the RF world of carriers and noise floors to the digital world of bits, error rates, and throughput.
Despite their fundamental importance, these metrics are routinely confused. Engineers mix up C/N with Eb/N0, forget bandwidth normalization when converting between them, or misread modem diagnostics that report Es/N0 instead of Eb/N0. This article provides a unified treatment: clear definitions, the conversion formulas, worked examples with real DVB-S2 parameters, and the common mistakes that cause errors in link budget analysis and system commissioning.
What Is C/N?
C/N stands for Carrier-to-Noise ratio. It is the ratio of the received carrier power to the total noise power within the measurement bandwidth, expressed in dB.
C/N (dB) = C (dBW) − N (dBW)Where:
- C — Total received carrier power at the demodulator input, in dBW.
- N — Total noise power within the noise bandwidth B, in dBW. Noise power is calculated as
N = k · Tsys · B, where k is Boltzmann's constant, Tsys is the system noise temperature, and B is the noise bandwidth in Hz.
C/N is what you see when you look at a carrier on a spectrum analyzer: the height of the carrier above the noise floor within the resolution bandwidth. It is also what your modem reports as "signal quality" or "SNR" during carrier acquisition and tracking.
The value of C/N depends on the EIRP of the transmitter, the path loss, the receiver's G/T, and critically, the bandwidth over which noise is measured. The same link will show different C/N values if measured in different bandwidths — which is exactly why C/N0 exists.
In a standard link budget, C/N is the bottom-line result:
C/N (dB) = EIRP + G/T − Lp − k − 10·log₁₀(B)Where Lp is path loss (dB), k is Boltzmann's constant (−228.6 dBW/K/Hz), and B is noise bandwidth (Hz).
What Is C/N0?
C/N0 stands for Carrier-to-Noise-Density ratio. It is the ratio of the received carrier power to the noise power spectral density (noise power per Hz), expressed in dB-Hz.
C/N0 (dB-Hz) = C (dBW) − N0 (dBW/Hz)Where:
- C — Total received carrier power, in dBW (same as in C/N).
- N0 — Noise power spectral density, in dBW/Hz. This equals
k · Tsys, where k is Boltzmann's constant and Tsys is the system noise temperature.
The key difference from C/N is that C/N0 normalizes noise to a 1 Hz bandwidth, removing the dependence on the actual measurement or occupied bandwidth. This makes C/N0 the natural metric for comparing links that operate at different bandwidths or data rates.
Relationship to C/N
The conversion between C/N and C/N0 is straightforward:
C/N0 (dB-Hz) = C/N (dB) + 10·log₁₀(B)Where B is the noise bandwidth in Hz. A carrier with C/N = 12 dB measured in a 36 MHz bandwidth has:
C/N0 = 12 + 10·log₁₀(36 × 10⁶) = 12 + 75.6 = 87.6 dB-HzEngineers use C/N0 when they need to compare signal quality across systems with different bandwidths — for example, when evaluating whether a satellite transponder can support a narrow 2 Msps carrier and a wide 45 Msps carrier on the same beam. The link budget for both carriers produces the same C/N0 (assuming the same EIRP density and G/T), even though their C/N values differ by more than 13 dB.
What Is Eb/N0?
Eb/N0 stands for Energy-per-Bit to Noise-Density ratio. It is the ratio of the energy contained in one information bit to the noise power spectral density, expressed in dB.
Eb/N0 (dB) = C/N0 (dB-Hz) − 10·log₁₀(Rb)Where:
- C/N0 — Carrier-to-noise-density ratio, in dB-Hz.
- Rb — Information bit rate in bits per second (bps).
Eb/N0 is the metric that directly determines bit error rate (BER) for a given modulation and coding combination. Every modulation scheme — QPSK, 8PSK, 16APSK, 32APSK — has a characteristic BER-vs-Eb/N0 curve, and every FEC code rate shifts that curve left (better coding gain) or right (less redundancy). The DVB-S2 standard defines required Eb/N0 thresholds for each MODCOD, and these thresholds are what ACM engines use to select the optimal operating point.
Why does digital communication care about energy per bit rather than total power? Because Eb/N0 normalizes away both the bandwidth and the data rate, isolating the fundamental efficiency of the modulation and coding scheme. Two systems with identical Eb/N0 will produce the same BER, regardless of whether one runs at 1 Mbps and the other at 100 Mbps.
There is a closely related metric: Es/N0 (Energy-per-Symbol to Noise-Density ratio), which many DVB-S2 modems report instead of Eb/N0. The relationship is:
Es/N0 (dB) = Eb/N0 (dB) + 10·log₁₀(η)Where η is the spectral efficiency in bits per symbol (equal to modulation order × code rate). For QPSK 3/4: η = 2 × 0.75 = 1.5 bits/symbol, so Es/N0 = Eb/N0 + 1.76 dB.
For a comprehensive look at how modulation order and code rate affect these thresholds, see the satellite modulation and coding guide.
How These Metrics Are Related
The three metrics form a chain of progressive normalization:
C/N0 (dB-Hz) = C/N (dB) + 10·log₁₀(B)
Eb/N0 (dB) = C/N0 (dB-Hz) − 10·log₁₀(Rb)
Eb/N0 (dB) = C/N (dB) + 10·log₁₀(B/Rb)Each step removes one system-dependent parameter:
| Metric | What It Normalizes Away | What It Still Depends On |
|---|---|---|
| C/N | Nothing — raw measurement | Bandwidth, data rate, modulation, coding |
| C/N0 | Bandwidth | Data rate, modulation, coding |
| Eb/N0 | Bandwidth and data rate | Modulation and coding only |
Intuition for the chain:
-
C/N → C/N0: You have a carrier sitting 10 dB above the noise floor in a 36 MHz transponder. To compare it against a carrier in a 72 MHz transponder, you normalize to 1 Hz — that is C/N0. The wider-bandwidth carrier has more total noise but the same noise density, so C/N0 is the fair comparison.
-
C/N0 → Eb/N0: Two links have the same C/N0, but one runs at 10 Mbps and the other at 100 Mbps. The faster link spreads the same carrier energy over ten times more bits, so each bit gets less energy. Eb/N0 captures this: it tells you how much energy is available per bit, which is what the demodulator and FEC decoder actually need to recover the data.
The practical consequence: C/N is what you measure, C/N0 is what you use for system comparison, and Eb/N0 is what you use to predict BER and select MODCODs.
Why These Metrics Matter in SATCOM
Modem Lock and Demodulation
A satellite modem needs a minimum C/N to acquire and lock onto a carrier. The acquisition threshold is typically 1–2 dB higher than the tracking threshold, because the modem must first detect the carrier's presence and estimate its frequency and timing before it can begin demodulation. If C/N drops below the tracking threshold, the modem loses lock and the link goes down.
Modern DVB-S2 modems display C/N (or Es/N0) in real time, giving operators an immediate view of link health. For more on how these modems integrate into the terminal RF chain, see satellite terminal architecture.
BER and Packet Error Performance
Once the modem is locked, Eb/N0 determines the bit error rate for the active MODCOD. Each modulation and coding combination has a "waterfall curve" — a plot of BER versus Eb/N0 that drops steeply from 10⁻² to 10⁻⁸ over a range of just 1–2 dB. Operating above the waterfall cliff gives essentially error-free performance; operating below it causes rapid degradation.
DVB-S2 defines quasi-error-free (QEF) performance at a packet error rate of 10⁻⁷, which corresponds to a specific Eb/N0 threshold for each MODCOD. These thresholds range from approximately 0.7 dB for QPSK 1/4 to approximately 12.7 dB for 32APSK 9/10.
Throughput and ACM
Adaptive Coding and Modulation (ACM) systems continuously measure C/N (or Es/N0) on the return channel and select the highest MODCOD that the current link conditions can support. When C/N is high (clear sky, beam center), the ACM engine selects higher-order modulation like 16APSK or 32APSK with high code rates, maximizing spectral efficiency and throughput. When C/N drops (rain fade, beam edge), the engine steps down to more robust MODCODs like QPSK 1/2, maintaining the link at reduced throughput.
The difference between the best and worst MODCODs in DVB-S2 is roughly a factor of 5 in spectral efficiency — so a link that fluctuates between 14 dB and 6 dB of C/N can see its throughput vary by 5:1. Understanding C/N and its relationship to MODCOD selection is essential for capacity planning on any ACM-enabled network.
Practical Examples
Example 1: DVB-S2 VSAT Downlink
A Ku-band GEO satellite delivers a DVB-S2 carrier to a 1.2 m VSAT terminal. Compute C/N, C/N0, and Eb/N0, and determine the MODCOD.
| Parameter | Value |
|---|---|
| Satellite downlink EIRP | 52.0 dBW |
| Free-space path loss (12.5 GHz, 36,000 km) | 205.8 dB |
| Atmospheric attenuation (clear sky) | 0.3 dB |
| Terminal G/T | 21.6 dB/K |
| Boltzmann's constant (k) | −228.6 dBW/K/Hz |
| Occupied bandwidth (B) | 36 MHz (75.6 dBHz) |
| Information bit rate (Rb) | 45 Mbps (76.5 dBHz) |
C/N0 = EIRP − Lp − La + G/T − k
C/N0 = 52.0 − 205.8 − 0.3 + 21.6 − (−228.6)
C/N0 = 96.1 dB-Hz
C/N = C/N0 − 10·log₁₀(B)
C/N = 96.1 − 75.6
C/N = 20.5 dB
Eb/N0 = C/N0 − 10·log₁₀(Rb)
Eb/N0 = 96.1 − 76.5
Eb/N0 = 19.6 dBWith an Eb/N0 of 19.6 dB (or equivalently, Es/N0 of approximately 15.8 dB for the selected MODCOD), the ACM engine can select 32APSK 9/10 — the highest MODCOD in DVB-S2 — delivering maximum spectral efficiency of approximately 4.45 bits/symbol.
Example 2: Doubling the Symbol Rate
Same link parameters, but the operator doubles the symbol rate to 72 Msps and increases the occupied bandwidth to 72 MHz to support a higher aggregate throughput of 90 Mbps.
C/N0 = 96.1 dB-Hz (unchanged — C/N0 does not depend on bandwidth)
C/N = 96.1 − 10·log₁₀(72 × 10⁶)
C/N = 96.1 − 78.6
C/N = 17.5 dB (dropped 3 dB from the previous 20.5 dB)
Eb/N0 = 96.1 − 10·log₁₀(90 × 10⁶)
Eb/N0 = 96.1 − 79.5
Eb/N0 = 16.6 dBThe C/N dropped by 3 dB because the noise bandwidth doubled while the carrier power stayed the same. C/N0 remained constant, confirming that it is the bandwidth-independent metric. The Eb/N0 is still high enough for 16APSK 3/4, but not for 32APSK 9/10 — the modem steps down one MODCOD level, accepting slightly lower spectral efficiency to accommodate the wider carrier.
Example 3: Rain Fade Event
The original 36 MHz carrier is operating in clear sky at C/N = 20.5 dB when a rain event introduces 4 dB of additional attenuation. Rain also increases the antenna noise temperature from 45 K to 130 K, degrading G/T by approximately 1.5 dB.
C/N (rain) = C/N (clear sky) − rain attenuation − G/T degradation
C/N (rain) = 20.5 − 4.0 − 1.5
C/N (rain) = 15.0 dB
Eb/N0 (rain) = C/N (rain) + 10·log₁₀(B/Rb)
Eb/N0 (rain) = 15.0 + 10·log₁₀(36×10⁶ / 45×10⁶)
Eb/N0 (rain) = 15.0 + (−0.97)
Eb/N0 (rain) = 14.0 dBThe 5.5 dB total degradation (4 dB attenuation + 1.5 dB noise increase) forces the ACM engine to step down from 32APSK 9/10 to approximately 16APSK 2/3, reducing throughput by roughly 40%. This is the dual effect of rain fade that many engineers underestimate: rain both attenuates the carrier and increases the noise temperature, hitting the C/N from both sides.
C/N vs Eb/N0 vs Link Margin
Engineers use these metrics for different purposes, and confusing them causes errors:
| Metric | What It Represents | Where It's Used | Units |
|---|---|---|---|
| C/N | Carrier power above noise floor in a given bandwidth | Spectrum analyzer readings, modem diagnostics, link budget bottom line | dB |
| C/N0 | Carrier power above noise density (1 Hz reference) | System comparison across different bandwidths, link budget intermediate step | dB-Hz |
| Eb/N0 | Energy per information bit above noise density | BER prediction, MODCOD selection threshold, modem FEC performance | dB |
| Link margin | Available metric minus required threshold | System design verification, fade margin analysis | dB |
Link margin is not a separate signal quality metric — it is the difference between the C/N (or Eb/N0) that the link actually achieves and the minimum C/N (or Eb/N0) that the demodulator requires to maintain the target BER. For example:
Link margin = C/N (available) − C/N (required)
Link margin = 20.5 dB − 15.0 dB = 5.5 dBThis 5.5 dB margin is the "budget" available to absorb rain fade, pointing errors, interference, and other degradations before the link fails. You can express margin using either C/N or Eb/N0 — the numerical result is the same because the bandwidth and bit rate terms cancel out in the subtraction.
The critical point: C/N values measured in different bandwidths cannot be directly compared without normalizing to C/N0 first. Similarly, Eb/N0 thresholds from different modulation schemes cannot be compared against C/N without applying the bandwidth and bit rate conversion.
Common Mistakes
-
Confusing C/N with Eb/N0. These are not interchangeable. A link with C/N = 12 dB may have Eb/N0 = 10 dB or Eb/N0 = 15 dB depending on the ratio of bandwidth to bit rate. Always apply the conversion formula.
-
Ignoring bandwidth when converting C/N to C/N0. C/N0 = C/N + 10·log₁₀(B). Forgetting the bandwidth term — or using the wrong bandwidth (occupied vs noise vs allocated) — produces errors of several dB.
-
Using information bit rate vs symbol rate inconsistently. Eb/N0 uses the information bit rate (after FEC encoding). Es/N0 uses the symbol rate. Mixing these up shifts the result by 10·log₁₀(code rate × bits per symbol), which can be 3–7 dB.
-
Misreading modem diagnostics: Es/N0 vs Eb/N0. Many DVB-S2 modems report Es/N0 (energy per symbol), not Eb/N0 (energy per bit). The difference depends on the active MODCOD: for QPSK 1/2, Es/N0 = Eb/N0 + 0 dB; for 16APSK 3/4, Es/N0 = Eb/N0 + 4.8 dB. Comparing Es/N0 readings against Eb/N0 threshold tables produces misleading results.
-
Forgetting that rain degrades C/N through both attenuation and increased noise temperature. Rain attenuates the carrier (reducing C) and raises the antenna noise temperature (increasing N). Both effects reduce C/N. At Ka-band, the noise temperature increase alone can cost 1–3 dB on top of the path attenuation. See the rain fade guide for detailed treatment.
-
Comparing C/N values measured in different bandwidths without normalizing. A C/N of 18 dB in a 36 MHz transponder is not the same link quality as 18 dB in a 72 MHz transponder. Always convert to C/N0 before comparing links with different bandwidths.
Frequently Asked Questions
What is a good C/N for satellite communication? It depends on the target MODCOD and application. For basic QPSK 1/2 operation (voice, low-rate data), a C/N of 4–5 dB in the occupied bandwidth is sufficient. For high-throughput 16APSK or 32APSK operation, C/N of 12–16 dB is needed. Most commercial Ku-band VSAT links are designed for clear-sky C/N of 12–20 dB, providing margin for rain fade and interference.
What does C/N0 mean? C/N0 is the carrier-to-noise-density ratio — the carrier power divided by the noise power in a 1 Hz bandwidth. It removes the bandwidth dependence from the signal quality measurement, making it the correct metric for comparing links that use different bandwidths. C/N0 is expressed in dB-Hz, with typical SATCOM values ranging from 70 to 100 dB-Hz depending on the link.
How is Eb/N0 related to BER? Every combination of modulation scheme and FEC code rate has a characteristic BER-vs-Eb/N0 curve. As Eb/N0 increases, BER drops — slowly at first, then very steeply through the "waterfall" region. DVB-S2 defines quasi-error-free (QEF) performance at PER = 10⁻⁷, and publishes the minimum Eb/N0 required for each MODCOD to achieve this threshold. Engineers design links to operate above these thresholds with adequate margin.
Why does rain fade reduce signal quality metrics? Rain has two effects. First, raindrops absorb and scatter the RF signal, reducing the carrier power C that reaches the receive antenna. Second, the rain itself is a source of thermal noise — warm raindrops radiate noise energy into the antenna, raising the system noise temperature Tsys. Both effects reduce C/N: lower C in the numerator and higher N in the denominator. At Ka-band, heavy rain can reduce C/N by 10 dB or more in tropical regions.
What is the difference between Es/N0 and Eb/N0?
Es/N0 is the energy per symbol divided by noise density. Eb/N0 is the energy per information bit divided by noise density. The relationship is Es/N0 = Eb/N0 + 10·log₁₀(η), where η is the spectral efficiency in bits per symbol (modulation order × code rate). DVB-S2 modems typically report Es/N0 because it is independent of the FEC code rate and more directly related to the demodulator's operating point.
Can I measure C/N0 directly?
Not with a single instrument reading. You measure C/N on a spectrum analyzer or modem, then convert: C/N0 = C/N + 10·log₁₀(B). Alternatively, you can compute C/N0 from the link budget using C/N0 = EIRP + G/T − Lp − k, which avoids any bandwidth term entirely. Some advanced carrier monitoring systems compute and display C/N0 directly from power spectral density measurements.
What C/N does a DVB-S2 modem need to lock? DVB-S2 modems can acquire and track carriers at C/N values as low as −2 to 0 dB (using QPSK 1/4 with pilot symbols). Practical acquisition thresholds are typically 1–2 dB above the minimum demodulation threshold for the selected MODCOD. For ACM operation, the modem locks at the lowest MODCOD and then steps up as C/N improves.
Why do different modems report different signal quality numbers? Modems may report C/N, Es/N0, or Eb/N0 — and the numerical values differ for the same link because they use different normalization. Additionally, some modems measure over the occupied bandwidth while others use the noise bandwidth, and some report post-equalizer SNR while others report pre-equalizer values. Always check the modem documentation to understand exactly which metric is displayed and how it is measured, particularly when comparing readings from different manufacturer equipment.
Key Takeaways
- C/N is the raw, measurable signal quality — the carrier power above the noise floor in a given bandwidth. It is what modems report and what spectrum analyzers display.
- C/N0 removes bandwidth dependence by normalizing noise to 1 Hz, making it the correct metric for comparing links with different bandwidths or aggregating multiple link segments.
- Eb/N0 removes both bandwidth and data rate dependence, isolating the fundamental performance of the modulation and coding scheme — it directly predicts BER.
- The conversion chain is: C/N → (+10·log₁₀B) → C/N0 → (−10·log₁₀Rb) → Eb/N0. Every step removes one system parameter.
- Rain fade hits C/N from both sides — attenuating the carrier and raising the noise temperature — making the total degradation worse than the path attenuation alone.
- Always verify which metric your modem reports (C/N, Es/N0, or Eb/N0) before comparing against threshold tables or making link budget decisions.
Related Articles
- Satellite Link Budget Calculation — Complete link budget guide where C/N is the bottom-line result of EIRP, G/T, and path loss
- Satellite Modulation and Coding Guide — How modulation order and FEC code rate determine the Eb/N0 threshold for each MODCOD
- Adaptive Coding and Modulation in Satellite Systems — How ACM engines use real-time C/N measurements to select the optimal MODCOD
- Rain Fade in Satellite Communications — How rain degrades C/N through both signal attenuation and increased noise temperature
- Satellite EIRP Explained — The transmit-side figure of merit that determines carrier power at the receiver
- Satellite G/T Explained — The receive-side figure of merit that determines noise performance and C/N
- Satellite Interference Explained — How C/I (carrier-to-interference) differs from C/N and affects overall link quality
Author
Categories
More Posts

Enterprise Satellite Internet: Use Cases, Architecture, and Vendor Selection
Comprehensive guide to enterprise satellite internet covering use cases, hybrid WAN architecture, performance SLAs, vendor selection criteria, and procurement best practices.

Satellite Link Budget Calculation | Complete Engineering Guide
Step-by-step guide to satellite link budget calculation covering EIRP, free-space path loss, G/T, Eb/No, fade margin, and real-world deployment scenarios for maritime, energy, and desert VSAT.

Hubless VSAT Networks Explained: When Satellite Terminals Communicate Without a Central Hub
Engineering guide to hubless VSAT networks covering mesh satellite architectures, terminal-to-terminal communication, latency advantages, trade-offs vs hub-and-spoke, and real-world deployment scenarios.
Newsletter
Join the community
Subscribe to our newsletter for the latest news and updates