Formula sheet

Communication Systems Formula Sheet

When computing the figure of merit of a DSB-SC receiver against an FM receiver in a GATE 2024 noise analysis problem, or dimensioning a GSM channel bandwidth during an RF system design exercise, the formulas for modulation indices, SNR, bandwidth, and probability of error become essential shortcuts. These formulas cover AM and FM analog modulation, noise in communication systems, digital modulation (ASK, FSK, PSK, QAM), pulse modulation, multiplexing, and information theory — covering the full GATE ECE and ISRO examinations for ECE and EI students.

ECE, EI

Amplitude Modulation (AM)

AM Signal Time Domain

s(t) = A_c[1 + \mu\cos(2\pi f_m t)]\cos(2\pi f_c t)

SymbolDescriptionUnit
A_cCarrier amplitudeV
\muModulation index (0 ≤ μ ≤ 1)dimensionless
f_cCarrier frequencyHz
f_mMessage frequencyHz

Worked example

Find modulation index if A_c = 10 V, A_m = 6 V.

Given: A_c = 10 V, A_m = 6 V

  1. μ = A_m / A_c
  2. μ = 6 / 10
  3. μ = 0.6

Answer: μ = 0.6 (60% modulation)

AM Power Distribution

P_T = P_c\left(1 + \frac{\mu^2}{2}\right)

SymbolDescriptionUnit
P_TTotal AM powerW
P_cCarrier powerW
\muModulation indexdimensionless

Worked example

AM transmitter with P_c = 1000 W, μ = 0.6. Find total power.

Given: P_c = 1000 W, μ = 0.6

  1. P_T = P_c × (1 + μ²/2)
  2. = 1000 × (1 + 0.36/2)
  3. = 1000 × (1 + 0.18)
  4. = 1000 × 1.18 = 1180 W

Answer: P_T = 1180 W (180 W in sidebands)

AM Bandwidth

BW_{AM} = 2f_m

SymbolDescriptionUnit
BW_{AM}AM bandwidthHz
f_mMaximum message frequencyHz

Worked example

AM radio with audio up to 5 kHz. Find channel bandwidth.

Given: f_m = 5 kHz

  1. BW = 2 × f_m
  2. = 2 × 5000
  3. = 10000 Hz

Answer: BW = 10 kHz

SSB Power Advantage

P_{SSB} = \frac{\mu^2 P_c}{4} = \frac{P_{USB}}{1}

SymbolDescriptionUnit
P_{SSB}Single sideband powerW

Worked example

Compare SSB power to AM for P_c = 1000 W, μ = 1.

Given: P_c = 1000 W, μ = 1

  1. AM total = P_c(1 + 1/2) = 1500 W
  2. Each sideband = P_c × μ²/4 = 1000/4 = 250 W
  3. SSB uses only 250 W to convey same info as 1500 W AM

Answer: SSB needs only 250 W vs 1500 W AM — 6× more efficient

Frequency Modulation (FM)

FM Modulation Index

\beta = \frac{\Delta f}{f_m} = \frac{k_f A_m}{f_m}

SymbolDescriptionUnit
\betaFM modulation indexdimensionless
\Delta fMaximum frequency deviationHz
f_mMessage frequencyHz
k_fFrequency sensitivityHz/V

Worked example

FM: Δf = 75 kHz, f_m = 15 kHz (commercial FM). Find β.

Given: Δf = 75 kHz, f_m = 15 kHz

  1. β = Δf / f_m
  2. = 75000 / 15000
  3. = 5

Answer: β = 5 (wideband FM)

Carson's Rule for FM Bandwidth

BW_{FM} \approx 2(\Delta f + f_m) = 2f_m(\beta + 1)

SymbolDescriptionUnit
BW_{FM}Approximate FM bandwidthHz

Worked example

Find commercial FM station bandwidth with β = 5, f_m = 15 kHz.

Given: β = 5, f_m = 15 kHz

  1. BW = 2 × f_m × (β + 1)
  2. = 2 × 15000 × (5 + 1)
  3. = 2 × 15000 × 6
  4. = 180000 Hz

Answer: BW ≈ 180 kHz (commercial FM stations are 200 kHz spaced)

FM SNR Improvement

\left(\frac{SNR_o}{SNR_i}\right)_{FM} = 3\beta^2(\beta + 1)

SymbolDescriptionUnit
SNR_oOutput SNRdimensionless
SNR_iInput SNR (baseband reference)dimensionless

Worked example

Compare FM figure of merit for β = 5 vs AM (DSBSC).

Given: β = 5 for FM; figure of merit for DSBSC = 1

  1. FM figure = 3β²(β+1) = 3×25×6 = 450
  2. DSBSC figure = 1
  3. FM improvement = 450/1 = 450 = 26.5 dB over DSBSC

Answer: FM outperforms DSBSC by 26.5 dB in SNR for β = 5

Noise in Communication Systems

Thermal Noise Power

N = kTB

SymbolDescriptionUnit
NThermal noise powerW
kBoltzmann constant = 1.38×10⁻²³ J/KJ/K
TTemperatureK
BBandwidthHz

Worked example

Find thermal noise power in a 1 MHz bandwidth at T = 290 K.

Given: k = 1.38e-23, T = 290, B = 1e6

  1. N = k × T × B
  2. = 1.38e-23 × 290 × 1e6
  3. = 1.38e-23 × 2.9e8
  4. = 4.002e-15 W

Answer: N ≈ 4 fW = -144 dBW = -114 dBm

Noise Figure

F = \frac{SNR_{in}}{SNR_{out}} = 1 + \frac{T_e}{T_0}

SymbolDescriptionUnit
FNoise figuredimensionless
T_eEquivalent noise temperatureK
T_0Reference temperature = 290 KK

Worked example

LNA with noise figure NF = 2 dB. Find equivalent noise temperature.

Given: NF = 2 dB → F = 10^(2/10) = 1.585

  1. F = 1 + T_e/T₀
  2. T_e = (F-1) × T₀
  3. = (1.585 - 1) × 290
  4. = 0.585 × 290
  5. = 169.7 K

Answer: T_e ≈ 170 K

Friis Noise Formula (Cascaded Stages)

F_{total} = F_1 + \frac{F_2 - 1}{G_1} + \frac{F_3 - 1}{G_1 G_2} + \ldots

SymbolDescriptionUnit
F_1, F_2Noise figures of stagesdimensionless
G_1, G_2Power gains of stagesdimensionless

Worked example

LNA: G1 = 20 dB, F1 = 1.585 (2 dB NF). Mixer: F2 = 10 (10 dB NF). Find F_total.

Given: G1 = 100 (linear), F1 = 1.585, F2 = 10

  1. F_total = F1 + (F2-1)/G1
  2. = 1.585 + (10-1)/100
  3. = 1.585 + 0.09
  4. = 1.675

Answer: F_total = 1.675 (2.24 dB) — dominated by the first stage

Digital Modulation

BPSK Probability of Error

P_e = Q\left(\sqrt{\frac{2E_b}{N_0}}\right)

SymbolDescriptionUnit
P_eBit error probabilitydimensionless
E_bEnergy per bitJ
N_0One-sided noise PSDW/Hz
Q(x)Q-function = P(Z>x) for Z~N(0,1)dimensionless

Worked example

Find P_e for BPSK at E_b/N_0 = 10 dB.

Given: E_b/N_0 = 10 dB = 10 (linear)

  1. Argument = √(2 × E_b/N_0) = √(2 × 10) = √20 = 4.47
  2. P_e = Q(4.47)
  3. Q(4.47) ≈ 3.9 × 10⁻⁶

Answer: P_e ≈ 3.9 × 10⁻⁶

QPSK Bandwidth Efficiency

\eta = \frac{R_b}{BW} = 2 \text{ bps/Hz (ideal)}

SymbolDescriptionUnit
\etaSpectral efficiencybps/Hz
R_bBit ratebps

Worked example

QPSK system with 10 MHz bandwidth. Find maximum bit rate.

Given: BW = 10 MHz, η = 2 bps/Hz

  1. R_b = η × BW
  2. = 2 × 10×10⁶
  3. = 20×10⁶ bps

Answer: R_b = 20 Mbps

QAM Spectral Efficiency

\eta = \log_2(M) \text{ bps/Hz},\quad M = \text{constellation size}

SymbolDescriptionUnit
MNumber of symbols (e.g., 16 for 16-QAM)dimensionless

Worked example

64-QAM in 10 MHz channel. Find bit rate.

Given: M = 64, BW = 10 MHz

  1. η = log₂(64) = 6 bps/Hz
  2. R_b = 6 × 10e6
  3. = 60e6 bps

Answer: R_b = 60 Mbps at 6 bps/Hz efficiency

FSK Minimum Bandwidth

BW_{FSK} = (M-1)\Delta f + 2B

SymbolDescriptionUnit
\Delta fFrequency spacingHz
MNumber of frequenciesdimensionless
BBaseband signal bandwidthHz

Worked example

Binary FSK (M=2): f1 = 1200 Hz, f2 = 2200 Hz, baud rate = 1200 baud. Find BW.

Given: Δf = 2200-1200 = 1000 Hz, B = 600 Hz (half baud rate)

  1. BW = (2-1)×1000 + 2×600
  2. = 1000 + 1200
  3. = 2200 Hz

Answer: BW ≈ 2.2 kHz

Information Theory

Shannon's Channel Capacity

C = B\log_2\left(1 + \frac{S}{N}\right)

SymbolDescriptionUnit
CChannel capacitybps
BBandwidthHz
S/NSignal-to-noise ratio (linear)dimensionless

Worked example

Find capacity of a 3 kHz voice channel with SNR = 1000 (30 dB).

Given: B = 3000 Hz, S/N = 1000

  1. C = 3000 × log₂(1 + 1000)
  2. = 3000 × log₂(1001)
  3. = 3000 × 9.967
  4. = 29902 bps

Answer: C ≈ 29.9 kbps

Entropy of a Source

H = -\sum_{i=1}^{M} p_i \log_2 p_i \text{ bits/symbol}

SymbolDescriptionUnit
HSource entropybits/symbol
p_iProbability of symbol idimensionless

Worked example

Binary source with p(0) = 0.25, p(1) = 0.75. Find entropy.

Given: p0 = 0.25, p1 = 0.75

  1. H = -(0.25 × log₂0.25 + 0.75 × log₂0.75)
  2. = -(0.25 × (-2) + 0.75 × (-0.415))
  3. = -((-0.5) + (-0.311))
  4. = -(-0.811) = 0.811 bits/symbol

Answer: H = 0.811 bits/symbol (max H = 1 bit for equal probability)

Nyquist Theorem (Noise-Free Channel Capacity)

C_{max} = 2B\log_2 M

SymbolDescriptionUnit
C_{max}Maximum noiseless channel capacitybps
MNumber of discrete signal levelsdimensionless

Worked example

Noiseless channel B = 3 kHz with 8 signal levels. Find max rate.

Given: B = 3000 Hz, M = 8

  1. C = 2 × 3000 × log₂(8)
  2. = 6000 × 3
  3. = 18000 bps

Answer: C_max = 18 kbps

Pulse Modulation and PCM

PCM Bit Rate

R_b = f_s \cdot n

SymbolDescriptionUnit
R_bPCM bit ratebps
f_sSampling frequencyHz
nBits per samplebits

Worked example

PCM for telephone: f_s = 8 kHz, 8 bits/sample. Find bit rate.

Given: f_s = 8000 Hz, n = 8

  1. R_b = f_s × n
  2. = 8000 × 8
  3. = 64000 bps

Answer: R_b = 64 kbps (standard G.711 PCM)

PCM Quantization Noise SNR

SNR_{PCM} \approx 6.02n + 1.76 \text{ dB}

SymbolDescriptionUnit
nNumber of bits per samplebits

Worked example

Find SNR for 8-bit PCM.

Given: n = 8

  1. SNR = 6.02 × 8 + 1.76
  2. = 48.16 + 1.76
  3. = 49.92 dB

Answer: SNR ≈ 50 dB for 8-bit PCM

TDM Frame Efficiency

\text{Efficiency} = \frac{n \cdot N}{n \cdot N + F} \times 100\%

SymbolDescriptionUnit
nBits per channel slotbits
NNumber of channelsdimensionless
FFraming bits per framebits

Worked example

TDM: 24 channels, 8 bits each, 1 framing bit per frame (T1 format). Find efficiency.

Given: N = 24, n = 8, F = 1

  1. Total bits per frame = n×N + F = 8×24 + 1 = 193
  2. Data bits = 192
  3. Efficiency = 192/193 × 100
  4. = 99.48%

Answer: Efficiency = 99.48% (T1 carrier)

Spread Spectrum and Multiple Access

Processing Gain (CDMA)

PG = \frac{BW_{spread}}{R_b} = \frac{f_c}{f_m}

SymbolDescriptionUnit
PGProcessing gaindimensionless
BW_{spread}Spread spectrum bandwidthHz
R_bData ratebps

Worked example

CDMA with chip rate 1.2288 Mcps and data rate 9.6 kbps (IS-95). Find PG.

Given: BW = 1.2288e6 Hz, R_b = 9600 bps

  1. PG = BW / R_b
  2. = 1228800 / 9600
  3. = 128

Answer: PG = 128 = 21 dB (IS-95 CDMA standard)

OFDM Number of Subcarriers

N = \frac{BW}{\Delta f}, \quad \Delta f = \frac{1}{T_s}

SymbolDescriptionUnit
NNumber of OFDM subcarriersdimensionless
\Delta fSubcarrier spacingHz
T_sOFDM symbol durations

Worked example

LTE 20 MHz BW with 15 kHz subcarrier spacing. Find N.

Given: BW = 20e6 Hz, Δf = 15000 Hz

  1. N = BW / Δf
  2. = 20000000 / 15000
  3. = 1333 subcarriers (LTE uses 1200 active + guard)

Answer: N ≈ 1333 total; 1200 active in LTE 20 MHz

Quick reference

FormulaExpression
AM Modulation Indexμ = A_m/A_c
AM Total PowerP_T = P_c(1 + μ²/2)
AM BandwidthBW = 2f_m
FM Modulation Indexβ = Δf/f_m
FM Bandwidth (Carson)BW = 2(Δf + f_m)
FM Figure of Merit3β²(β+1)
Thermal NoiseN = kTB
Noise FigureF = SNR_in/SNR_out
Friis FormulaF = F1 + (F2-1)/G1 + ...
Shannon CapacityC = B log₂(1 + S/N)
Source EntropyH = -Σ p_i log₂ p_i
BPSK P_eP_e = Q(√(2E_b/N_0))
QAM Efficiencyη = log₂(M) bps/Hz
PCM Bit RateR_b = f_s × n
PCM SNRSNR = 6.02n + 1.76 dB

Exam tips

  • GATE noise analysis problems specify whether SNR is pre-detection or post-detection — DSB-SC and SSB have figure of merit 1 while FM's figure 3β²(β+1) can far exceed this.
  • For AM efficiency questions, efficiency = μ²/(2+μ²) × 100% — at 100% modulation (μ=1) only 33.3% of total power is in the sidebands (useful).
  • Shannon capacity is an upper bound — actual systems need coding; examiners ask you to compare achieved bit rate to capacity to find spectral efficiency.
  • In digital modulation BER problems, always convert E_b/N_0 from dB to linear before substituting into Q-function formulas.
  • Friis noise formula shows that the first stage dominates total system noise — a high-gain low-noise first amplifier is critical, tested in receiver design problems.
  • For PCM, each additional bit increases SNR by approximately 6 dB — GATE frequently asks how many bits are needed to achieve a target SNR.