Since heated dust has been determined to be pervasive in galaxies, then the aggregate of all the galaxies in the Universe should produce a redshift-smeared background over the range 10-400 µ. In general, the spectrum of dust emission is that of a blackbody convolved with the emissivity of the dust grains. The total energy emitted by the dust scales as T_dustⁿ⁺⁴ where n is the emissivity index, ⁿ. Large grains, which dominate the emission at long wavelengths, have an emissivity which goes as ^-1 and hence the total energy goes as T⁵. This is an important point of energy conservation and balance in galaxies. Small differences in dust temperature between galaxies or between regions in the same galaxy, reflect very large differences in energy input. Simple modeling of the interaction between the radiation field and the dust temperature (see Bothun, Lonsdale and Rice 1989) suggests that the dust temperature is a good diagnostic of the nature of the heating sources (e.g., UV radiation from newly formed stars vs. the ambient light from older stars). Since the extragalactic background represents the sum of sources at different wavelengths, then it will not be characterized by a simple blackbody of some given temperature. However, its possible that there might be a "feature" in the spectrum that would represent high star formation rates at high redshift. This, of course, assumes that dust, which must come from previous generations of massive star formation, is already in place in these galaxies at high redshift.

A detection of a possible cosmological infrared background (hereafter the CIB), means detecting an isotropic signal, of unknown spectral signature, against the strong signal of our Galaxy, which, due to high latitude cirrus emission, is nearly isotropic itself. Further difficulty arises when trying to calibrate the absolute strength of the CIB as all the strong foreground sources need to be properly removed. To unambiguously detect the CIB then requires very good modeling of the the known solar system and Galactic foregrounds to examine if there are significant sources of residual emission that are isotropically distributed. Attempts to do this with the IRAS data did not yield any strong results. However, one of the instruments on board COBE was the Diffuse Infrared Background Experiment (DIRBE) which made measurements over the range 1.25-240 µ (significantly longer than IRAS). The DIRBE measurements are potentially quite sensitive to the existence of the CIB.

Estimated minimum strengths of this background can be obtained by using the existing far-infrared (FIR) Luminosity function (LF) of galaxies, as determined by IRAS. Over the range 100-300 µ, these minimum strengths are in the range 2-4 x 10^-9 Watts m^-2 sr^-1 when the FIR LF is integrated out to z = 3. Any luminosity evolution in the sources and/or increase in space density with redshift would make the real background potentially much higher than this minimum estimate. The current status of the modeling, coupled with the DIRBE observations, shows residuals of 20-50 x 10^-9 Watts m^-2 sr^-1 at high latitude in this wavelength range.

While this is a possible detection of the CIB, at a level at least 10 times the predicted minimum strength, it could equally as well be an indication that the current modeling of foreground sources is inadequate. Further, FIR missions, such as ISO and SIRTF will help to better determine the CIB if it exists. Moreover, even if the CIB were unambiguously established its origin would still be difficult to interpret. The low angular resolution at FIR means that many discrete sources (e.g., galaxies) would fill a single beam. The expectation, of course, is that the CIB is produced by galaxies, but at high redshift, intergalactic dust heated by QSOs could also produce a CIB (see Heisler and Ostriker 1988).