Figures of merit & link margin

budget.summary()

LinkContainer.summary() computes the budget and returns a BudgetSummary dataclass. Print it for a readable report, or read the fields directly:

summary = budget.summary()
print(summary)
print(summary.ebno_db, summary.g_over_t_db, summary.closes)
data = summary.as_dict()          # plain dict, stages included

Always populated

Field	Meaning
signal_power_w, signal_power_dbw	final signal power
noise_power_w, noise_power_dbw	final noise power
snr, snr_db	final signal-to-noise ratio

Populated when there is enough context

Field	Needs	Meaning
cn_db	noise_bandwidth	carrier-to-noise ratio in that bandwidth
cn0_dbhz	noise_bandwidth	carrier-to-noise density (dB-Hz)
ebno_db	noise_bandwidth + data_rate	energy-per-bit to noise density
esno_db	noise_bandwidth + symbol_rate	energy-per-symbol to noise density
shannon_capacity_bps	noise_bandwidth	Shannon–Hartley capacity
spectral_efficiency_bps_per_hz	noise_bandwidth + data_rate	data_rate / noise_bandwidth
eirp_dbw	a propagation stage	signal power entering the first one
total_propagation_loss_db	a propagation stage	product of every propagation stage's loss
system_noise_temp_k	noise_bandwidth + a noise-floor stage	Friis sum of every stage's noise, referred to the floor plane
g_over_t_db	the above + a role="rx" antenna	receiver figure of merit
margin_db, closes	a LinkMargin stage	see below

System noise temperature and the noise model

system_noise_temp_k sums each stage's own added noise, referred to the noise-floor plane and divided down by the gain ahead of it. A stage only contributes if it reports a thermal contribution: a ThermalNoise floor, an AntennaNoiseTemperature, or a noise-figure stage running the Friis model (noise_bandwidth set). Noise-figure stages left on the default multiplicative model, and quantisation noise, don't add a temperature — with only those present the value falls back to the floor-plane noise power expressed as a temperature.

LinkMargin

Add a LinkMargin as the last stage. It passes signal and noise through untouched and annotates the stage (and hence the summary) with the closure verdict.

Give it exactly one requirement:

Argument	Requirement is on	Also needs
required_ebno_db	E​b/N₀	noise_bandwidth, data_rate
required_esno_db	E​s/N₀	noise_bandwidth, symbol_rate
required_cn_db	C/N	noise_bandwidth
required_snr_db	raw SNR	—

budget.add_component(linkbudget.LinkMargin(
    "Link margin", "QPSK rate-1/2",
    required_ebno_db=1.0,
    noise_bandwidth=5e6,
    data_rate=4e6,
    implementation_loss_db=1.5,   # subtracted from the achieved metric
    coding_gain_db=0.0,           # added to the achieved metric
))

The margin is achieved − required + coding_gain − implementation_loss; closes is margin_db >= 0.

The linkbudget.fom module

The figure-of-merit maths lives in pure functions you can use on their own:

from linkbudget import fom

fom.eirp_dbw(tx_power_dbw=0.0, tx_antenna_gain_dbi=36.0)          # 36.0
fom.free_space_path_loss_db(distance_m=12e3, frequency_hz=6e9)     # ~129.6
fom.g_over_t_db(rx_antenna_gain_dbi=36.0, system_noise_temp_k=290) # ~11.4
fom.cn0_from_snr(snr_db=12.0, noise_bandwidth_hz=20e6)             # ~85.0
fom.ebno_db(cn0=85.0, data_rate_bps=45e6)                          # ~8.5

See the linkbudget.fom API reference for the full set.