What is DSP?
Key takeaways Digital signal processing (DSP) is manipulating a signal as a stream of numbers using arithmetic, instead of with analog circuits. Filtering, tuning, and demodulation all become math a computer runs. DSP is the engine inside a software-defined radio: the hardware just digitizes the airwaves, and DSP turns those numbers into voice, data, and control messages.
This is lesson 1 of the DSP path. Its job is to answer what DSP is and why it matters before we get into the how. By the end you’ll see where every later lesson — sampling, filters, the FFT, demodulation — fits in the chain.
From circuits to numbers
For most of radio’s history, a signal was processed by physical circuits: a coil and capacitor tuned to a station, a diode recovered the audio. Each function needed dedicated hardware. DSP replaces those circuits with arithmetic. Once a signal is a list of numbers, “tune to this frequency” or “keep only these components” is a calculation, and one general-purpose processor can do the work of a rack of analog gear.
That shift is the whole idea behind software-defined radio: the radio hardware does as little as possible — just convert the antenna’s signal into numbers — and software does the rest. Change the software and the same hardware decodes a different system.
What a signal looks like as numbers
A signal is just a value that changes over time — a voltage, a pressure, a radio wave’s amplitude. Sample it regularly and you get a sequence:
time: 0 1 2 3 4 5 6 ...
value: 0.0 0.6 0.9 0.6 0.0 -0.6 -0.9 ... (a sampled sine wave)
Everything DSP does operates on sequences like this. Smoothing noise is averaging neighbouring values; tuning is multiplying by another sequence; detecting a tone is asking how much of a given frequency is present. The next lesson, sampling & quantization, covers exactly how a continuous wave becomes this list.
Where DSP sits in a software radio
Everything between “digitize” and “decoded data” is DSP, and it’s what this module builds up stage by stage. By the final lesson you’ll map each stage onto GopherTrunk’s real code.
Why numbers win
Processing in software buys you three things analog can’t match easily:
- Flexibility — a new protocol is new code, not new hardware.
- Precision — arithmetic is exact and repeatable; components drift with heat and age.
- Complexity for free — filters and detectors that would need dozens of parts are a few lines of math.
The price is that you need enough compute to keep up with the sample rate — which is why performance is a real concern in DSP code.
Quick check: what does DSP operate on?
Recap
- DSP manipulates a signal as a stream of numbers with arithmetic, replacing analog circuits.
- It’s the engine of software-defined radio: hardware digitizes, software decodes.
- Every DSP operation is math on a sequence of samples.
- Numbers buy flexibility, precision, and complexity — at the cost of needing compute.
Next up: how a continuous wave actually becomes that stream of numbers.
Frequently asked questions
What is digital signal processing in simple terms?
Digital signal processing is doing to a signal — with arithmetic on numbers — what used to be done with analog electronic circuits. Once a signal is a stream of numbers, filtering, tuning, and decoding all become math a computer performs. That flexibility is what makes software-defined radio possible.
Do I need to be good at maths to learn DSP?
Not to start. The core ideas are intuitive — averaging numbers to smooth a signal, sliding one list along another, looking at a signal by its frequencies. This module keeps the maths light and visual; you can understand the whole pipeline with arithmetic and pictures, and pick up the deeper formulas later if you want them.