Modules 1–4 built the analog world: a message rides a carrier by varying its amplitude (AM) or its frequency (FM), and the receiver reads that continuous variation back. Every modern wireless system — WiFi, LTE, 5G, Bluetooth — does something different first: it turns the message into bits, and modulates those. This module is about digital modulation, and it opens with the obvious question: if analog worked for a century, why did the whole world switch?
An analog signal carries information in a continuously varying quantity — the exact voltage, at every instant, is the message. A digital signal carries information as a sequence of discrete symbols, almost always bits (0s and 1s). The message is first sampled and quantized into numbers, and it is those numbers, encoded as symbols, that the carrier transmits.
Intuition: Analog is a dimmer knob; digital is a light switch with many labelled positions. A dimmer knocked slightly off is slightly wrong forever. A switch nudged between positions still snaps to the nearest label — the small error vanishes. That snapping-back is the single idea behind almost every advantage below.
Noise adds a little error to every received signal. In analog, that error is permanent and accumulates at every relay. In digital, the receiver only has to decide which symbol was sent — and as long as the noise is smaller than half the gap between symbols, it decides correctly and regenerates a perfect copy. A digital signal can be relayed across a continent through hundreds of hops and arrive bit-for-bit identical; an analog signal degrades a little at each one.
Because the message is numbers, you can add redundant numbers computed from them — a checksum, a parity bit, or a full error-correcting code. The receiver recomputes and compares, detecting and often repairing corrupted bits with no retransmission. There is no analog equivalent: you cannot “checksum” a continuously varying voltage.
Numbers can be compressed. Removing redundancy and perceptually irrelevant detail (as JPEG, MP3, and H.264 do) lets far more content fit in the same bandwidth. Analog signals cannot be compressed in any comparable way.
Bit streams from many users interleave cleanly onto one channel and separate again perfectly at the far end — in time, in frequency, or by code. This is what lets one cell tower serve thousands of phones at once, the subject of Module 10. Mixing analog signals leaves them permanently entangled.
Bits can be encrypted with mathematically provable strength; a continuous waveform cannot. Every secure wireless link on Earth is digital for this reason alone.
All of this raises a hard question: how many bits per second can a channel actually carry? In 1948 Claude Shannon answered it exactly. The channel capacity — the absolute maximum error-free data rate — depends only on the bandwidth B and the signal-to-noise ratio:
Shannon’s result is the reason digital modulation is not just convenient but optimal: it tells us a hard ceiling exists, and every scheme in this module — ASK, FSK, PSK, QAM — is an attempt to pack bits toward that ceiling. The trade always comes down to bits per symbol against robustness to noise:
Digital is not free. Converting analog to digital needs an analog-to-digital converter and, done naively, more bandwidth than the original analog signal. It also has a harsher failure mode: analog degrades gracefully into static, while digital works perfectly until it suddenly does not — the cliff effect, familiar from digital TV that freezes and blocks rather than fading to snow. The rest of Module 5 is about spending bandwidth wisely to stay well clear of that cliff.