Digital Transmission

Digital transmission is the process of sending data using digital signals. It forms the foundation of modern networking, enabling reliable communication between computers, servers, routers, smartphones, and other digital devices.

What Is Digital Transmission?

Digital transmission is the process of transmitting information using discrete digital signals, typically represented as binary values (0s and 1s). Unlike analog transmission, which uses a continuously varying signal, digital transmission uses a small set of specific voltage (or light, or radio) levels to represent binary data. Because only a few discrete levels are valid, a receiver can tell exactly which value was intended even after the signal has picked up some noise along the way — it just needs to decide which of the few valid levels is the closest match.

Real-World Example

When you send a WhatsApp message:

1. The text is converted into binary data (0s and 1s).
2. The binary data is encoded into a digital signal.
3. The signal travels through network cables, fiber optics, or wireless channels.
4. The receiving device decodes the signal and reconstructs the original message.

This entire process depends on digital transmission.

Why Is Digital Transmission Important?

Digital transmission offers several advantages over analog transmission:

• Higher noise immunity — small amounts of noise don't change which discrete level the receiver detects
• Better data accuracy — errors can often be caught before they corrupt the meaning of the data
• Easier error detection and correction — techniques like checksums and CRC work naturally on discrete values
• Improved security through encryption — digital data can be mathematically transformed in ways analog signals cannot
• Compatibility with modern computers and digital devices, which store and process everything as binary

Because of these advantages, virtually all modern communication systems — from Ethernet to Wi-Fi to fiber-optic backbones — use digital transmission.

Digital-to-Digital Conversion

Digital-to-digital conversion is the process of converting digital data (the bits a computer generates) into a digital signal suitable for transmission over a communication medium.

A computer produces binary digits (0s and 1s), but bits by themselves are just abstract values — they cannot travel through a cable on their own. They must first be encoded into a physical signal: a pattern of electrical voltage, light intensity, or electromagnetic wave. This encoding process is called digital-to-digital encoding (or line coding).

Digital encoding techniques fall into three broad categories:

1. Unipolar Encoding
2. Polar Encoding
3. Bipolar Encoding

1. Unipolar Encoding

Unipolar encoding uses only one polarity — positive voltage — to represent binary data.

Example

Where +V = positive voltage and 0 = no voltage.

Advantages

• Very simple to implement
• Low cost

Disadvantages

DC component problem. Because only positive voltage is ever used, the average signal level over time is not zero — this creates a DC component that most transmission media and equipment handle poorly (it wastes power and doesn't transfer well over media designed to carry alternating signals).

Synchronization problem. Long runs of the same bit value produce no voltage transitions at all, which can cause the receiver to lose track of exactly where one bit ends and the next begins.

Because of these limitations, unipolar encoding is rarely used in modern communication systems.

Understanding synchronization. Synchronization means the sender and receiver stay aligned on the timing of each transmitted bit — both sides must agree on exactly when one bit period ends and the next begins. Suppose the sender transmits 11111111. If the signal never changes, the receiver has no timing reference and may lose count of how many 1s were actually sent. Good encoding schemes deliberately introduce signal transitions to give the receiver something to "clock" against, which is why most practical line codes are built around guaranteeing regular transitions.

2. Polar Encoding

Polar encoding uses two voltage levels: positive voltage (+V) and negative voltage (−V). Because the signal spends time at both positive and negative levels, the average voltage tends toward zero, which reduces the DC component problem seen in unipolar encoding.

NRZ (Non-Return-to-Zero)

NRZ is one of the simplest polar encoding techniques. The signal holds a constant level for the entire bit period (it does not return to zero in the middle of the bit, hence the name). There are two common forms.

NRZ-L (Non-Return-to-Zero Level)

The voltage level directly represents the bit value:

NRZ-I (Non-Return-to-Zero Inverted)

In NRZ-I, the bit value is represented by whether the voltage changes, not by the absolute level:

For binary data 10110, the signal inverts its polarity every time a 1 occurs, and stays the same for every 0.

Advantages of NRZ-I: better synchronization than NRZ-L (every 1 forces a transition), and less sensitive to accidental polarity inversion of the whole signal (e.g., if the two wires get swapped, the pattern of transitions is still readable).

Disadvantage: long runs of 0s still produce no transitions, so synchronization problems remain for that case.

RZ (Return-to-Zero)

In Return-to-Zero encoding, the signal returns to zero voltage in the middle of every bit period, rather than only at bit boundaries. This uses three voltage levels: positive, negative, and zero.

Advantage: because every single bit period contains a transition back to zero, synchronization is much more reliable than with NRZ.

Disadvantages: RZ requires more bandwidth than NRZ, since up to two signal transitions can occur per bit instead of one.

Biphase Encoding

Biphase encoding guarantees a transition in the middle of every bit period, which makes synchronization highly reliable without needing a separate zero level. The two common forms are Manchester Encoding and Differential Manchester Encoding.

Manchester Encoding

Manchester encoding forces a transition at the center of every bit:

Advantages: excellent synchronization, and the signal is "self-clocking" — the receiver can recover timing directly from the data transitions without a separate clock signal.

Disadvantage: requires roughly twice the bandwidth of NRZ for the same data rate, since there are up to two transitions per bit.

Real-world use: traditional 10 Mbps Ethernet (10BASE-T and its predecessors) used Manchester encoding.

Differential Manchester Encoding

Differential Manchester also guarantees a transition in the middle of every bit for synchronization, but it identifies the bit value differently — through whether a transition occurs at the start of the bit period, not through the direction of the mid-bit transition:

Advantages: excellent synchronization, and — because the bit value depends on the presence/absence of a transition rather than absolute polarity — it is more resistant to having the two signal wires accidentally reversed.

Applications: used in various communication and data storage systems, including early Token Ring networks.

3. Bipolar Encoding

Bipolar encoding uses three voltage levels: positive (+V), negative (−V), and zero (0).

For binary data 1111, the signal alternates: +V −V +V −V. Because consecutive 1s alternate polarity, the average voltage stays near zero, eliminating the DC component problem.

AMI (Alternate Mark Inversion)

AMI is the most common bipolar encoding technique. The term "mark" is a historical telegraphy term meaning binary 1 (and "space" means binary 0).

Advantages: no DC component, and good synchronization as long as there's a steady stream of 1s (each 1 produces a transition).

Disadvantage: long sequences of 0s still produce no voltage transitions at all, so synchronization problems persist for that case — which is exactly the problem the next two techniques were designed to fix.

B8ZS (Bipolar 8-Zero Substitution)

Long runs of zeros generate no signal transitions, and without transitions the receiver can lose synchronization. B8ZS solves this by replacing every run of eight consecutive zeros with a special substitute pattern that deliberately violates the normal alternating-polarity rule of AMI. The receiver recognizes this "bipolar violation" pattern, knows it represents eight zeros, and restores the original data — while still getting the transitions it needs to stay synchronized.

Features: developed in North America; based on AMI encoding; maintains synchronization even during long runs of zeros.

Applications: widely used in T1 digital transmission systems (the standard 1.544 Mbps digital carrier used in North American telecom networks).

HDB3 (High-Density Bipolar 3)

HDB3 solves the same long-zero-run problem as B8ZS, but with a different substitution rule: instead of waiting for eight consecutive zeros, HDB3 substitutes after every run of four consecutive zeros.

Features: developed in Europe and Japan; improves synchronization; eliminates long zero sequences from the transmitted signal.

Applications: widely used in E-carrier telecommunications systems (the European equivalent of the T-carrier system).

Analog-to-Digital Conversion

Many real-world signals are naturally analog — human voice, music, temperature readings, and video signals are all continuously varying quantities. To process, store, or transmit these signals digitally, they must first be converted into digital form. This process is called Analog-to-Digital Conversion (ADC).

Why Convert Analog Signals to Digital?

Digital signals offer:

• Better resistance to noise
• Easier storage
• More reliable transmission
• More efficient processing (compression, error correction, encryption all work naturally on digital data)

Real-World Example

During a phone call:

1. Your voice is an analog sound wave.
2. The smartphone's microphone and ADC circuitry sample the voice at regular intervals.
3. Those samples are converted into digital data.
4. The digital data is packaged and transmitted through the network.
5. The receiver's device reconstructs an approximation of the original voice from the digital data.

Techniques for Analog-to-Digital Conversion

The two major techniques are:

1. PAM (Pulse Amplitude Modulation)
2. PCM (Pulse Code Modulation)

PAM (Pulse Amplitude Modulation)

Pulse Amplitude Modulation is the first step in converting an analog signal into digital form — it is the sampling stage.

How PAM works

1. Measure (sample) the analog signal's amplitude at regular time intervals.
2. Create a series of pulses whose amplitudes correspond to the measured values at each interval.

This measuring step is called sampling. For the resulting samples to accurately represent the original signal, they must be taken often enough — a well-known result in signal processing (the Nyquist sampling theorem) states that the sampling rate must be at least twice the highest frequency present in the original signal, or information will be lost.

Example

Imagine measuring the height of ocean waves once every second. Each measurement becomes one pulse — a snapshot of the wave's height at that instant.

Limitation

PAM produces a series of pulses, but each pulse's amplitude can still take any value within a continuous range — the amplitude itself hasn't been converted to a fixed set of binary numbers yet. Because of this, PAM by itself does not produce a true digital signal; it's an intermediate step on the way to one.

PCM (Pulse Code Modulation)

PCM is the most widely used analog-to-digital conversion technique. It takes the sampling idea from PAM and carries it all the way through to fully digital binary data.

The PCM Process

PCM consists of four major stages:

1. Sampling — measure the analog signal's amplitude at fixed, regular intervals (as in PAM).
2. Quantization — round each sample to the nearest one of a fixed, finite set of predefined amplitude levels. This is the step that actually makes the signal digital: instead of an unlimited range of possible values, each sample now takes one of a limited set of levels.
3. Binary encoding — convert each quantized level into a binary number.
4. Digital transmission — transmit the resulting binary data using one of the digital-to-digital encoding techniques covered earlier (such as NRZ, AMI, or Manchester).

Practical Applications of PCM

PCM underlies a wide range of everyday digital audio systems, including:

• Digital telephony
• Voice over IP (VoIP)
• Audio CDs
• Digital audio recording
• Video conferencing systems