Glossary of Communication Terms

Glossary of Communication Systems Terms

This glossary covers essential terminology you'll encounter throughout your communication systems course. Each term includes context explaining why it matters and how it connects to other concepts. Understanding the relationships between terms is just as valuable as knowing individual definitions.

A

Amplitude Modulation (AM): A modulation technique where the message signal varies the amplitude of a high-frequency carrier wave. AM is the simplest modulation scheme but wastes power — at least 67% goes to the carrier which carries no information. Used in commercial AM radio broadcasting (530-1700 kHz).

Analog-to-Digital Converter (ADC): Hardware that converts continuous analog signals into discrete digital values. Critical parameters: sampling rate (must satisfy Nyquist criterion fs ≥ 2fm) and resolution (bits per sample, determining dynamic range ≈ 6 dB per bit).

Antenna Gain: Ratio of power radiated in a specific direction to what an isotropic antenna would radiate. Measured in dBi. A 20 dBi antenna focuses energy 100× more but only in one direction. Appears in link budget as Gt and Gr.

AWGN (Additive White Gaussian Noise): The standard channel model. "Additive" means noise adds to signal. "White" means flat power spectral density N₀/2 at all frequencies. "Gaussian" means amplitude follows normal distribution. Provides the theoretical baseline for BER analysis.

B

Bandwidth: The range of frequencies occupied by a signal or supported by a channel. For baseband (0 to fm): BW = fm. For passband (f1 to f2): BW = f2 - f1. Shannon shows capacity is directly proportional to bandwidth — it's the most precious communication resource.

Baseband: The original frequency range of a signal before modulation. Voice (300-3400 Hz), digital data streams are baseband signals. Baseband transmission (Ethernet, USB) sends signals directly without shifting to higher frequencies.

BER (Bit Error Rate): Probability that a received bit differs from transmitted. BPSK: BER = Q(√(2Eb/N₀)). BER of 10⁻⁶ = one error per million bits. Improves (decreases) as signal power increases or noise decreases.

Beamforming: Using multiple antennas to focus energy toward a specific direction through phase adjustment. Constructive interference in desired direction, destructive elsewhere. Used in 5G massive MIMO with 64-256 elements.

C

Carrier: High-frequency sinusoid that "carries" information. Information is embedded by varying amplitude (AM), frequency (FM), or phase (PM). Carrier frequency determines propagation characteristics and antenna size.

Channel Capacity: Maximum error-free transmission rate: C = B·log₂(1 + SNR) bits/s. This is Shannon's fundamental limit — no practical system can exceed it, though modern codes (LDPC, Polar) come within 0.1 dB.

CDMA (Code Division Multiple Access): Multiple users share the same frequency simultaneously using unique spreading codes. Each user's signal appears as noise to others. A correlator extracts the desired signal. Used in 3G cellular and GPS.

Coherent Detection: Demodulation requiring carrier phase knowledge. Receiver multiplies received signal by local carrier replica. Provides 3 dB advantage over non-coherent detection but needs carrier recovery (PLL, Costas loop).

Constellation Diagram: Graphical representation showing all possible symbol points in I-Q plane. QPSK has 4 points, 16-QAM has 16 in a grid. Distance between nearest points determines noise immunity.

D-F

Demodulation: Extracting original message from modulated carrier. AM uses envelope detection or coherent multiplication. FM uses frequency discriminator. Digital uses matched filtering plus threshold decision.

Diversity: Combating fading by receiving multiple independent copies. Types: spatial (multiple antennas), frequency (multiple bands), time (multiple transmissions), polarization. If one copy fades deeply, others likely don't.

Eb/N₀ (Energy per Bit to Noise Density): The fundamental digital communication metric. Normalizes SNR to per-bit basis for fair comparison across modulation schemes. Relationship: Eb/N₀ = SNR × (B/Rb).

Entropy: Information content measure: H(X) = -Σ p(x)·log₂p(x) bits/symbol. Maximum when all outcomes equally probable. A fair coin has H = 1 bit; a biased coin has H < 1 bit. Source coding aims for code length approaching entropy.

Fading: Signal strength variation from multipath. Rayleigh fading: no line-of-sight, deep nulls possible (30-40 dB below average). Rician fading: dominant direct path plus reflections, less severe. Fast fading changes within symbols; slow fading persists over many.

FDMA: Each user gets a dedicated frequency band continuously. Simplest multiple access (used in 1G cellular). Wasteful when user is silent — bandwidth sits idle.

FFT (Fast Fourier Transform): Efficient DFT algorithm: O(N·log₂N) vs O(N²). Essential for OFDM — converts between time and frequency domains. A 2048-point FFT in LTE executes every 71.4 μs.

G-N

Guard Interval: Time gap between symbols preventing ISI from multipath. In OFDM, implemented as cyclic prefix (copy of symbol end prepended). Must exceed maximum delay spread τmax. Wastes efficiency by Tg/(Ts + Tg).

ISI (Intersymbol Interference): One symbol's energy spilling into adjacent periods from channel dispersion. Nyquist criterion: ISI-free transmission needs minimum BW = Rs/2 Hz. Raised-cosine filtering satisfies this practically.

Link Budget: Accounting all gains/losses: Pr = Pt + Gt + Gr - FSPL - Lfading - Lmargins. Received power must exceed sensitivity by fade margin (10-30 dB) for acceptable BER during fluctuations.

MIMO: Multiple antennas at both ends. Spatial multiplexing: independent streams on different antennas (throughput × min(Nt,Nr)). Spatial diversity: same data for reliability. Capacity: C = B·log₂det(I + (SNR/Nt)·H·H†).

Modulation Index: How much carrier is modified. AM: μ = Am/Ac (keep ≤ 1). FM: β = Δf/fm. Higher index means more information capacity but wider transmitted bandwidth.

Nyquist Rate: Minimum sampling rate for perfect reconstruction: fs = 2fm. Below this causes aliasing — irreversible corruption where high frequencies fold into low frequencies.

O-S

OFDM: Multicarrier modulation dividing wideband channel into narrow subcarriers. Each experiences flat fading (easy to equalize). Subcarriers overlap but remain orthogonal (no interference). Used in WiFi, LTE, 5G, DVB-T.

Path Loss: Signal power reduction with distance. Free-space: FSPL = (4πd/λ)². In dB: 20·log₁₀(d) + 20·log₁₀(f) + 32.44 (d in km, f in MHz). Real environments add shadowing and multipath.

Q-Function: Q(x) = P(standard normal > x). Core of all digital BER expressions. Q(3) ≈ 1.3×10⁻³, Q(4) ≈ 3.2×10⁻⁵, Q(5) ≈ 2.9×10⁻⁷. Small SNR changes → large BER changes.

Rayleigh Fading: No line-of-sight model. Envelope: p(r) = (r/σ²)·exp(-r²/(2σ²)). Creates deep nulls (30-40 dB drops) causing burst errors. Diversity techniques combat this.

Spread Spectrum: Transmitting over much wider bandwidth than necessary. Direct Sequence multiplies by PN code; Frequency Hopping rapidly changes frequency. Provides interference rejection and multiple access.

Symbol Rate (Baud): Symbol changes per second. Symbol rate = Bit rate / log₂(M). A 64-QAM at 1 MBaud = 6 Mbps. Bandwidth depends on symbol rate, not bit rate.

T-W

TDMA: Users share frequency via different time slots. Each gets periodic full-bandwidth access. Used in GSM (8 users per 200 kHz). Requires tight synchronization.

Thermal Noise: Random electron motion noise: N = kTB. Unavoidable. Sets fundamental receiver sensitivity. At room temperature: noise floor = -174 dBm/Hz. Also called Johnson-Nyquist noise.

Viterbi Algorithm: Maximum-likelihood sequence detection for systems with memory (convolutional codes, ISI). Explores trellis states but prunes unlikely paths — complexity linear in sequence length, not exponential.

Wavelength: λ = c/f. At 900 MHz (GSM): 33 cm. At 5 GHz (WiFi): 6 cm. At 28 GHz (5G mmWave): 1.07 cm. Determines antenna size (typically λ/4), path loss, and diffraction around obstacles.

Key Takeaways

1. Every term connects to others — bandwidth determines capacity, capacity determines bit rate, bit rate determines BER requirements, and BER determines required Eb/N₀.
2. The three fundamental resources are power, bandwidth, and time — every system design trades between them.
3. Understanding units (linear vs dB, Hz vs bps, watts vs dBm) is essential for solving numerical problems correctly.
4. Shannon's capacity theorem connects information theory to physical reality — every other formula either approaches this limit or explains why systems fall short.
5. Modern systems (5G, WiFi 6) combine multiple techniques — MIMO + OFDM + LDPC + beamforming — each term describing one piece of a sophisticated whole.
6. Focus on why each technique exists (what problem it solves) rather than just memorizing definitions — this deeper understanding helps in both exams and engineering practice.