Electromagnetic Waves, Antennas, and RF: Thermal Noise, Noise Temperature, Noise Figure,
G∕T, and C∕N0 - Exercises and Complete Worked Solutions
This companion to EM26 turns receiver-noise theory into calculation practice. The exercises begin
with Johnson–Nyquist noise and kTB, move through equivalent noise temperature and noise
figure, and then build complete cascaded receiver calculations. The final problems combine system
noise temperature, G∕T, C∕N0, and bandwidth-dependent C∕N in RF and GNSS-style examples
[1, 2, 3, 4, 5].
The principal relations are
and
Use
Unless otherwise stated, gains and losses used in cascade formulas are power ratios in linear
units.
Figure. White thermal-noise density integrates over receiver bandwidth. The density is
N0 = kT, while the total available noise power is N = kTB.
Part I: Exercises
Exercise 1: Johnson noise voltage and matched noise power
A 50.0 Ω resistor is at 290 K and is observed over a 1.00 MHz bandwidth. Find (a) the RMS
open-circuit Johnson-noise voltage, (b) the available noise power delivered to a matched load, and
(c) that noise power in dBm. Verify that the matched power is independent of the resistance
value.
Exercise 2: thermal-noise power in a receiver bandwidth
At 290 K, calculate the available thermal-noise power in a 2.00 MHz bandwidth. Express the result
in watts, dBW, and dBm. Compare it with the familiar −174 dBm/Hz room-temperature noise
density.
Exercise 3: noise density at a colder temperature
Find N0 = kT at T = 150 K in (a) W/Hz, (b) dBW/Hz, and (c) dBm/Hz. By how many decibels
is this below the 290 K noise density?
Figure. Equivalent input noise temperature increases nonlinearly with noise figure. The
conversion uses Te = (10NF∕10 − 1)T
0.
Exercise 4: noise figure to equivalent noise temperature
An amplifier has noise figure NF = 1.20 dB. Find its linear noise factor F and equivalent input
noise temperature Te referred to T0 = 290 K.
Exercise 5: equivalent noise temperature to noise figure
A receiver front end has equivalent input noise temperature Te = 80.0 K. Find its noise factor and
noise figure.
Exercise 6: three-stage cascaded receiver
A three-stage receiver has
| G1 | = 20.0 dB, | NF1 | = 1.00 dB, | (8)
|
| G2 | = 15.0 dB, | NF2 | = 4.00 dB, | (9)
|
| G3 | = 10.0 dB, | NF3 | = 6.00 dB. | (10) |
Calculate the total input-referred noise factor, total noise figure, and total equivalent input
noise temperature. Quantify how strongly the first-stage gain suppresses stages 2 and
3.
Figure. Friis cascade noise formula. High gain in the first low-noise stage reduces the
input-referred importance of later-stage noise.
Exercise 7: why loss before an LNA is expensive
A cable has 1.50 dB power loss at 290 K. An LNA has 1.00 dB noise figure and 20.0 dB gain.
Compare two cases: (a) cable before LNA and (b) cable after LNA. Find the total cascade noise
figure and equivalent input noise temperature in both cases.
Exercise 8: antenna temperature and system noise temperature
An antenna has noise temperature TA = 120 K and a receiver referred to the same input plane has
equivalent noise temperature Te,rx = 80.0 K. Find (a) Tsys, (b) N0 in W/Hz, and (c) N0 in
dBW/Hz and dBm/Hz.
Exercise 9: receiver figure of merit G∕T
A receiving antenna has directional gain 28.0 dBi and the complete receiver system has Tsys = 150
K at the same reference plane. Calculate G∕T in dB/K. Explain what physical change would
improve G∕T by 3 dB.
Exercise 10: C∕N0 from carrier power and system temperature
At the receiver reference plane, the carrier power is C = −155 dBW and Tsys = 250 K. Calculate
the noise density in dBW/Hz and the resulting C∕N0 in dB-Hz.
Exercise 11: direct C∕N0 link budget using G∕T
An illustrative RF link has
Use the direct link-budget form to find C∕N0. Use −10 log 10k ≈ 228.60 dB-K/Hz.
Figure. C∕N0 compares carrier power with noise density. Choosing a bandwidth integrates
that density and converts the result to C∕N.
Exercise 12: converting C∕N0 to C∕N
A receiver measures C∕N0 = 45.0 dB-Hz. What is C∕N in a noise-equivalent bandwidth of 2.00
MHz?
Exercise 13: infer receiver bandwidth from C∕N0 and C∕N
A receiver has C∕N0 = 48.0 dB-Hz and measured C∕N = 8.00 dB. Assuming thermal
noise dominates and the stated bandwidth is a noise-equivalent bandwidth, determine
B.
Exercise 14: bandwidth penalty
A fixed carrier has C∕N0 = 50.0 dB-Hz. Compute C∕N for bandwidths of 10.0 kHz and 20.0
kHz. Show explicitly that doubling bandwidth degrades C∕N by approximately 3.01
dB.
Exercise 15: GNSS-style weak-signal example
An illustrative GNSS-like receiver has carrier power C = −158.5 dBW and system noise
temperature Tsys = 400 K. Find (a) N0 in dBW/Hz, (b) C∕N0 in dB-Hz, and (c) C∕N in a 2.00
MHz front-end noise bandwidth. Interpret how C∕N can be negative while C∕N0 remains a useful
receiver-quality metric.
Exercise 16: reference-plane bookkeeping with feed loss and an LNA
At the antenna terminals, TA = 100 K. A 1.00 dB feed loss at 290 K precedes an LNA having
NF = 0.80 dB and 25.0 dB gain. Refer all noise to the antenna-terminal plane. Find (a) the
feed equivalent noise temperature, (b) the LNA equivalent input noise temperature, (c)
total system noise temperature, and (d) G∕T if the receive antenna gain is 25.0 dBi. If
the carrier at the same antenna-terminal reference plane is C = −160 dBW, also find
C∕N0.
Part II: Complete Worked Solutions
Solution 1: Johnson noise voltage and matched noise power
The open-circuit mean-square voltage is
Therefore
| vn,rms | =  | (13)
|
| = 8.95 × 10−7 V. | (14) |
Thus
With a matched load, only half the open-circuit voltage appears across the load. The delivered
power is
| N | =  | (16)
|
| =  | (17)
|
| = kTB. | (18) |
Hence
| N | = (1.380649 × 10−23)(290)(1.00 × 106) | (19)
|
| = 4.0039 × 10−15 W. | (20) |
In dBm,
| NdBm | = 10 log 10 | (21)
|
| = −113.98 dBm. | (22) |
Therefore
The resistance cancels algebraically. Changing R changes the noise voltage, but a matched
resistor at the same temperature and bandwidth has the same available noise power
kTB.
Solution 2: thermal-noise power in a receiver bandwidth
Using
we obtain
| N | = (1.380649 × 10−23)(290)(2.00 × 106) | (25)
|
| = 8.0078 × 10−15 W. | (26) |
Thus
The logarithmic values are
| NdBW | = 10 log 10(8.0078 × 10−15) = −140.96 dBW, | (28)
|
| NdBm | = −110.96 dBm. | (29) |
At 290 K the density is approximately −173.98 dBm/Hz. A 2.00 MHz bandwidth contributes
so
in agreement with the direct calculation.
Solution 3: noise density at a colder temperature
At 150 K,
| N0 | = kT | (32)
|
| = (1.380649 × 10−23)(150) | (33)
|
| = 2.0710 × 10−21 W/Hz. | (34) |
Therefore
| N0,dBW/Hz | = 10 log 10(2.0710 × 10−21) = −206.84 dBW/Hz, | (35)
|
| N0,dBm/Hz | = −176.84 dBm/Hz. | (36) |
Thus
Relative to 290 K, the decrease is
10 log 10 | = −2.86 dB. | (38) |
So cooling from 290 K to 150 K lowers thermal-noise density by about 2.86 dB.
Solution 4: noise figure to equivalent noise temperature
Convert noise figure to linear noise factor:
| F | = 10NF∕10 | (39)
|
| = 101.20∕10 | (40)
|
| = 1.3183. | (41) |
Then
| Te | = (F − 1)T0 | (42)
|
| = (1.3183 − 1)(290) | (43)
|
| = 92.3 K. | (44) |
Therefore
Solution 5: equivalent noise temperature to noise figure
Starting from
we find
| F | = 1 +  | (47)
|
| = 1.27586. | (48) |
Therefore
| NF | = 10 log 10(1.27586) | (49)
|
| = 1.058 dB. | (50) |
Hence
Solution 6: three-stage cascaded receiver
First convert all gains and noise figures to linear ratios:
| G1 | = 1020∕10 = 100, | (52)
|
| G2 | = 1015∕10 = 31.6228, | (53)
|
| G3 | = 1010∕10 = 10, | (54) |
and
| F1 | = 101∕10 = 1.25893, | (55)
|
| F2 | = 104∕10 = 2.51189, | (56)
|
| F3 | = 106∕10 = 3.98107. | (57) |
The cascade noise factor is
| Ftot | = F1 + +  | (58)
|
| = 1.25893 + +  | (59)
|
| = 1.27499. | (60) |
Thus
| NFtot | = 10 log 10(1.27499) | (61)
|
| = 1.055 dB. | (62) |
The equivalent input temperature is
| Te,tot | = (Ftot − 1)T0 | (63)
|
| = (0.27499)(290) | (64)
|
| = 79.7 K. | (65) |
Therefore
The stage-2 contribution is only
and stage 3 contributes only
This quantifies how the 20 dB first-stage gain suppresses later noise when referred back to the
receiver input.
Solution 7: why loss before an LNA is expensive
A 1.50 dB cable loss corresponds to
so the cable power gain is
At 290 K, the cable noise factor equals its loss:
The LNA has
For cable first,
| Fbefore | = Fc +  | (73)
|
| = 1.41254 +  | (74)
|
| = 1.77828. | (75) |
Hence
The equivalent input temperature is
For LNA first,
| Fafter | = FL +  | (78)
|
| = 1.25893 +  | (79)
|
| = 1.26305. | (80) |
Thus
and
The same cable is far more damaging before the LNA because its loss directly attenuates the
desired signal and its own noise is not suppressed by preceding gain.
Solution 8: antenna temperature and system noise temperature
At a common reference plane,
| Tsys | = TA + Te,rx | (83)
|
| = 120 + 80 | (84)
|
| = 200 K. | (85) |
Therefore
| N0 | = kTsys | (86)
|
| = (1.380649 × 10−23)(200) | (87)
|
| = 2.7613 × 10−21 W/Hz. | (88) |
The logarithmic forms are
| N0,dBW/Hz | = −205.59 dBW/Hz, | (89)
|
| N0,dBm/Hz | = −175.59 dBm/Hz. | (90) |
Hence
Solution 9: receiver figure of merit G∕T
The figure of merit is
dB/K | = 28.0 − 10 log 10(150) | (92)
|
| = 28.0 − 21.761 | (93)
|
| = 6.239 dB/K. | (94) |
Thus
A 3 dB improvement could come from approximately doubling receive gain at unchanged system
temperature, or halving system temperature at unchanged gain. Either changes the linear ratio
G∕T by a factor of two.
Solution 10: C∕N0 from carrier power and system temperature
At Tsys = 250 K,
| N0 | = kTsys | (96)
|
| = 3.4516 × 10−21 W/Hz. | (97) |
In dBW/Hz,
Therefore
dB-Hz | = CdBW − N0,dBW/Hz | (99)
|
| = −155 − (−204.62) | (100)
|
| = 49.62 dB-Hz. | (101) |
Thus
Solution 11: direct C∕N0 link budget using G∕T
Use
Substitution gives
| C∕N0 | = 27.0 − 182.5 − 2.0 − 20.0 + 228.60 | (104)
|
| = 51.10 dB-Hz. | (105) |
Therefore
This form combines transmitter strength, path loss, miscellaneous losses, and receiver sensitivity
into one bandwidth-independent link metric.
Solution 12: converting C∕N0 to C∕N
For B = 2.00 MHz,
| 10 log 10B | = 10 log 10(2.00 × 106) | (107)
|
| = 63.010 dB-Hz. | (108) |
Thus
| C∕N | = 45.0 − 63.010 | (109)
|
| = −18.01 dB. | (110) |
Therefore
The negative value means that the integrated noise power over the full 2 MHz bandwidth exceeds
the carrier power. It does not imply that the signal is necessarily unusable by a receiver employing
correlation or other signal-processing gain.
Solution 13: infer receiver bandwidth from C∕N0 and C∕N
From
we obtain
Therefore
| B | = 1040∕10 | (114)
|
| = 104 Hz. | (115) |
Thus
Solution 14: bandwidth penalty
For B1 = 10.0 kHz,
| C∕N1 | = 50.0 − 10 log 10(104) | (117)
|
| = 50.0 − 40.0 | (118)
|
| = 10.0 dB. | (119) |
For B2 = 20.0 kHz,
| C∕N2 | = 50.0 − 10 log 10(2.00 × 104) | (120)
|
| = 50.0 − 43.010 | (121)
|
| = 6.990 dB. | (122) |
Hence
Doubling bandwidth doubles integrated white-noise power, and a factor of two in power is 3.010
dB.
Solution 15: GNSS-style weak-signal example
For Tsys = 400 K,
| N0,dBW/Hz | = 10 log 10(kTsys) | (124)
|
| = −202.58 dBW/Hz. | (125) |
Therefore
| C∕N0 | = −158.5 − (−202.58) | (126)
|
| = 44.08 dB-Hz. | (127) |
For B = 2.00 MHz,
| C∕N | = 44.08 − 63.010 | (128)
|
| = −18.93 dB. | (129) |
Thus
C∕N0 compares a finite carrier power with noise per unit bandwidth. The wideband C∕N
includes all noise integrated across 2 MHz. Spread-spectrum and correlation receivers can
exploit known signal structure even when the wideband pre-correlation C∕N is below 0
dB.
Solution 16: reference-plane bookkeeping with feed loss and an LNA
The 1.00 dB feed loss is
with gain
At physical temperature 290 K, the feed equivalent input temperature is
| Te,c | = (L − 1)Tp | (133)
|
| = (1.25893 − 1)(290) | (134)
|
| = 75.09 K. | (135) |
The LNA noise factor is
so
| Te,L | = (FL − 1)T0 | (137)
|
| = 58.66 K. | (138) |
Referred through the preceding cable to the antenna-terminal plane, the LNA contribution
is
Therefore
| Tsys | = TA + Te,c +  | (140)
|
| = 100 + 75.09 + 73.85 | (141)
|
| = 248.93 K. | (142) |
Thus
For receive gain Gr = 25.0 dBi,
| G∕T | = 25.0 − 10 log 10(248.93) | (144)
|
| = 1.039 dB/K. | (145) |
Therefore
Finally,
| N0,dBW/Hz | = 10 log 10(kTsys) | (147)
|
| ≈−204.64 dBW/Hz. | (148) |
With C = −160 dBW at the same reference plane,
| C∕N0 | = −160 − (−204.64) | (149)
|
| = 44.64 dB-Hz. | (150) |
Hence
This problem illustrates why reference-plane discipline matters: the feed noise, the attenuated LNA
noise contribution, the receive gain, and carrier power must all be referred consistently before
forming Tsys, G∕T, or C∕N0.
Julia numerical check
The following script checks the room-temperature noise floor, the three-stage cascade, bandwidth
scaling, and the GNSS-style C∕N0 example.
using Printf
k = 1.380649e-23
T0 = 290.0
# Thermal noise density
N0 = k*T0
@printf("N0 at 290 K = %.4e W/Hz = %.3f dBm/Hz\n",
N0, 10*log10(N0/1e-3))
# Three-stage cascade
G1, G2 = 10^(20/10), 10^(15/10)
F1, F2, F3 = 10^(1/10), 10^(4/10), 10^(6/10)
Ftot = F1 + (F2-1)/G1 + (F3-1)/(G1*G2)
@printf("Cascade NF = %.4f dB\n", 10*log10(Ftot))
# C/N from C/N0 versus bandwidth
cno = 50.0
for B in (1e4, 2e4, 1e5, 1e6)
cn = cno - 10*log10(B)
@printf("B=%9.0f Hz C/N=%8.3f dB\n", B, cn)
end
# GNSS-style example
C = -158.5
Tsys = 400.0
N0_dBW_Hz = 10*log10(k*Tsys)
cno_gnss = C - N0_dBW_Hz
@printf("GNSS-style C/N0 = %.3f dB-Hz\n", cno_gnss)
The expected outputs are approximately −173.98 dBm/Hz at 290 K, 1.055 dB cascade noise figure,
a 3.010 dB reduction in C∕N when bandwidth doubles from 10 to 20 kHz, and 44.08 dB-Hz for the
illustrative GNSS-style case.
What EM26E1 adds to the series
EM26 introduced the thermal-noise quantities. EM26E1 makes their reference planes and
logarithmic conversions operational. The calculation chain is
The next article can now add received interference power J and derive J∕S, J∕N0, and the
degradation of carrier quality in simultaneous thermal noise and interference.
References
[1] J. B. Johnson, “Thermal Agitation of Electricity in Conductors,” Physical Review,
vol. 32, pp. 97–109, 1928.
[2] H. Nyquist, “Thermal Agitation of Electric Charge in Conductors,” Physical Review,
vol. 32, pp. 110–113, 1928.
[3] D. M. Pozar, Microwave Engineering, 4th ed., Wiley, 2012.
[4] C. A. Balanis, Antenna Theory: Analysis and Design, 4th ed., Wiley, 2016.
[5] B. Sklar, Digital Communications: Fundamentals and Applications, 2nd ed., Prentice
Hall, 2001.