The Dark Ages represent the last frontier in cosmology, the era
between the genesis of the cosmic microwave background (CMB) at
recombination and the formation of the first stars. During the Dark
Ages, when the Universe was unlit by any star, the only detectable
signal is likely to be that from neutral hydrogen (HI), which will
appear in absorption against the CMB. The HI absorption represents
potentially the richest of all data sets in cosmology—not only is the
underlying physics relatively simple so that the Hi absorption can be
used to constrain fundamental cosmological parameters in a manner
similar to that of CMB observations, but the spectral nature of the
signal allows the evolution of the Universe as a function of redshift
(z) to be followed. The Hi absorption occurs in dark matter-dominated
overdensities, locations that will later become the birthplaces of the
first stars, so tracing this evolution will provide crucial insights
into the properties of dark matter and potentially reveal aspects of
cosmic inflation. Moreover, given the relatively simple physics—the
Universal expansion, Compton scattering between CMB photons and
residual electrons, and gravity—any deviation from the expected
evolution would be a “clean” signature of fundamentally new physics.

The Dark Ages Lunar Interferometer (DALI) is a mission proposed for
study to NASA for a telescope located on the far side of the Moon, the
only site in the solar system shielded from human-generated
interference and, at night, from solar radio emissions. The DALI array
will observe at 3–30 m wavelengths (10–100 MHz; redshifts 15 \le z \le
150), and the DALI baseline concept builds on ground-based telescopes
operating at similar wavelengths, e.g., the Long Wavelength Array
(LWA) and Murchison Widefield Array (MWA). Specifically, the
fundamental collecting element will be dipoles. The dipoles will be
grouped into “stations,” deployed via rovers over an area of
approximately 50 km in diameter to obtain the requisite angular
resolution. The desired three-dimensional imaging requires
approximately 1000 stations, each containing 100 dipoles (i.e., ~ 105
dipoles); alternate processing approaches may produce useful results
with significantly fewer dipoles (factor ~ 3–10). Each station would
be deployed by one rover, which would also serve as a “transmission
hub” for sending the signals for correlation to a central processing
facility. After sending the correlator output to Earth, analysis would
then proceed via standard methods being developed for ground-based
 arrays.