Introduction to Digital Communication

Introduction to Digital Communication

We live in a digital world. Every phone call you make, every video you stream, every message you send — the underlying information travels as sequences of zeros and ones. But the physical world is analog — sound waves, light, and radio waves are continuous. Digital communication is the art and science of representing continuous information as discrete digital signals and transmitting them reliably across imperfect channels.

Why Digital? The Revolution Explained

To understand why the world moved from analog to digital communication, imagine making a photocopy of a photocopy of a photocopy — each generation gets worse. That is analog communication in a nutshell. Every time an analog signal is amplified or relayed, the noise accumulated along the way gets amplified too, degrading quality irreversibly.

Now imagine retyping a document at each relay point. As long as you can read the original correctly, your retyped version is perfect — no degradation accumulates. This is the digital advantage: at each repeater, the signal is regenerated (not just amplified), eliminating accumulated noise.

Key advantages of digital over analog:

1. Noise immunity — Digital signals can be regenerated perfectly at repeater stations
2. Error correction — Redundancy bits allow receivers to detect and correct transmission errors
3. Encryption — Digital data can be easily encrypted for secure communication
4. Multiplexing — Time-division multiplexing allows flexible sharing of channel capacity
5. Integration — Voice, video, and data can share the same digital infrastructure
6. Storage — Digital signals can be stored indefinitely without quality loss
7. Processing — Digital signal processing enables adaptive equalization, filtering, and compression

The Sampling Theorem: Bridge Between Analog and Digital

The foundation of digital communication is the Nyquist-Shannon Sampling Theorem:

A bandlimited signal with maximum frequency fm can be perfectly reconstructed from samples taken at rate fs ≥ 2fm.

The minimum sampling rate 2fm is called the Nyquist rate.

Think of it this way: if you take snapshots of a pendulum swinging at 1 Hz, you need at least 2 snapshots per second to determine its frequency. If you sample too slowly (below Nyquist rate), you get aliasing — the reconstructed signal appears to have a different frequency than the original.

Example: Telephone speech bandwidth = 4 kHz → Nyquist rate = 8 kHz → 8000 samples per second. This is exactly the standard used in telephone systems worldwide (G.711 PCM).

Architecture of a Digital Communication System

A complete digital communication system consists of these functional blocks:

Transmitter Side:

1. Source — Generates the information (voice, video, data)
2. Source Encoder — Removes redundancy, compresses data (e.g., MP3, H.264)
3. Channel Encoder — Adds controlled redundancy for error protection (e.g., convolutional codes, LDPC)
4. Digital Modulator — Maps bits to analog waveforms for transmission (e.g., QPSK, 16-QAM)
5. Transmitter Front-End — Frequency upconversion, amplification, antenna

Channel — The physical medium (air, fiber, copper) that introduces noise, fading, and distortion

Receiver Side:

1. Receiver Front-End — Antenna, low-noise amplifier, downconversion
2. Digital Demodulator — Recovers digital symbols from noisy analog waveform
3. Channel Decoder — Detects and corrects errors using redundancy
4. Source Decoder — Decompresses data back to original format
5. Sink — Delivers information to user (speaker, display, storage)

Quantization: Converting Samples to Numbers

After sampling, each sample value must be represented by a finite number of bits — this is quantization. With n bits per sample, we can represent 2ⁿ distinct levels:

• 8 bits → 256 levels (telephone quality)
• 16 bits → 65,536 levels (CD quality)
• 24 bits → 16.7 million levels (studio quality)

Quantization error is the difference between the actual sample value and its quantized representation. For uniform quantization:

Signal-to-Quantization-Noise Ratio: SQNR = 6.02n + 1.76 dB

Each additional bit improves SQNR by approximately 6 dB. This rule of thumb is fundamental to digital audio and communication system design.

Bit Rate and Bandwidth

The bit rate of a digitized signal equals:

Rb = fs × n bits/second

Where fs is sampling rate and n is bits per sample.

Example: CD audio:

• Sampling rate: 44.1 kHz
• Bits per sample: 16
• Channels: 2 (stereo)
• Bit rate: 44,100 × 16 × 2 = 1.41 Mbps

The bandwidth required to transmit a digital signal depends on the modulation scheme used. For binary modulation with rectangular pulses:

Minimum bandwidth = Rb/2 Hz (Nyquist bandwidth)

In practice, practical filters require more bandwidth: typically 1 to 2 times Rb for binary modulation.

Digital Modulation: Mapping Bits to Waves

Digital modulation converts bit sequences into analog waveforms suitable for transmission. The three fundamental parameters of a carrier that can be varied are:

• Amplitude → ASK (Amplitude Shift Keying)
• Frequency → FSK (Frequency Shift Keying)
• Phase → PSK (Phase Shift Keying)

Modern systems combine amplitude and phase modulation to create QAM (Quadrature Amplitude Modulation), achieving high spectral efficiency.

The choice of modulation involves trade-offs:

• Higher-order modulation (more bits per symbol) → better bandwidth efficiency but needs higher SNR
• Lower-order modulation → more robust but uses more bandwidth

Performance Metrics

Digital communication system performance is measured by:

1. Bit Error Rate (BER) — Probability of a received bit being wrong (typical target: 10⁻⁶ to 10⁻⁹)
2. Spectral Efficiency — Bits per second per Hertz of bandwidth (bits/s/Hz)
3. Power Efficiency — Required Eb/N₀ to achieve target BER
4. Latency — End-to-end delay from source to sink
5. Throughput — Actual useful data rate after removing overhead

The Shannon Limit

Claude Shannon proved that every channel has a maximum capacity:

C = B × log₂(1 + SNR)

No communication system can transmit reliably faster than this limit, but arbitrarily low error rates are achievable at any rate below capacity with sufficiently sophisticated coding. This theorem guides all modern communication system design.

Key Takeaways

1. Digital communication represents continuous signals as discrete bit sequences, enabling noise-free regeneration and error correction.
2. The Nyquist sampling theorem requires sampling at least twice the highest frequency to avoid aliasing.
3. Quantization introduces controllable error — each additional bit provides 6 dB improvement in signal quality.
4. The system architecture separates source coding (compression), channel coding (error protection), and modulation (physical transmission).
5. BER, spectral efficiency, and power efficiency are the key metrics balancing reliability against resource usage.
6. Shannon capacity sets the ultimate speed limit that no practical system can exceed.