This paper reviews how astronomers plan to measure the large-scale distribution of neutral hydrogen using 21‑centimetre intensity mapping. The 21‑cm line is radio emission from neutral hydrogen that traces matter in the universe. Instead of detecting individual galaxies, intensity mapping records the combined emission across the sky and through frequency, which lets telescopes probe large volumes quickly. The Square Kilometre Array Mid telescope, SKA‑Mid, aims to use this single‑dish or “autocorrelation” mode to reach redshifts up to about z ≈ 3 and map hydrogen across cosmic time.

The authors summarize recent advances in the steps that turn raw radio measurements into a usable 21‑cm map. That pipeline includes flagging radio frequency interference (RFI), calibrating the instrument, forming maps from each dish, and separating the faint cosmological signal from much brighter foregrounds. They emphasize work done in controlled simulations. Typical simulated data cubes include bright foregrounds (Galactic synchrotron and free‑free emission and extragalactic radio sources), a model 21‑cm brightness temperature up to z ≈ 3, a simplified telescope beam, and instrument noise. The SKA‑Mid AA4 array with 197 dishes is expected to give lower noise than a smaller 144‑dish array in these studies.

A central problem is that foregrounds are enormously brighter than the 21‑cm signal. Galactic synchrotron emission can be about five orders of magnitude stronger. Extragalactic point sources can be roughly three orders of magnitude stronger. The chapter reviews methods to remove those foregrounds. Blind and non‑parametric methods, which do not assume a detailed sky model, have proved powerful in simulations. The authors also discuss alternative strategies such as multi‑frequency angular power spectrum approaches, Bayesian methods that work in higher dimensional spaces, and machine learning (AI) driven solutions.