Astroengineering is engineering at astronomical scale, i.e. at planetary, stellar, stellar system, galactic or even larger scale. It is a form of megascale engineering. An example is the hypothetical Dyson sphere.
Typical megastructures that could hypothetically be produced by astroengineering include Dyson spheres, ringworlds, Alderson disks, Banks' Orbitals, Matrioshka brains, stellar engines such as Shkadov thrusters, and other artifacts produced by Type II and Type III civilizations on the Kardashev scale.
In a 2005 paper, Luc Arnold proposed a means of detecting smaller, though still megascale, artifacts from their distinctive transit light curve signature.
Some applications are becoming more relevant to our time period as advances in technology progress. Astroengineering projects like orbital towers (or orbital elevators) are becoming more technically feasible due to such technological advances.
It was suggested that an advanced civilization could use a pulsar for signalling through the modulation of its light curve. Such an application would require construction of a large scale structure (modulator) around the pulsar. Possible observable effects are missing pulses and an increase in thermal radiation due to the secondary emission from the modulator. Missing pulses (pulse nulling) has been observed in many pulsars and is usually attributed to natural phenomena. However, a statistical study of the missing pulse sequences revealed that nulling is not random in many pulsars. This poses a problem of distinguishing the natural modulation from artificial one.
Changing the cycle parameters of variable stars (Cepheids) is theoretically possible by firing a high-energy neutrino beam into the star's core. This effect can be used as a way to transmit information and create "galactic Internet". The rate of transmission is very slow, though, about 180 bits per year.