Annu. Rev. Astron. Astrophys. 2005. 43: 727-768 Copyright © 2005 by Annual Reviews. All rights reserved

1. INTRODUCTION

The Cosmic Infrared Background (CIB) can be defined as the part of the present radiation content of the Universe that is made essentially of the long wavelength output from all sources throughout the history of the Universe. The radiation content in the microwave part of the spectrum is dominated by the Cosmic Microwave Background (CMB) produced in the hot and dense phases of the universe. It dominates for frequencies below 800 GHz. Nevertheless the very different spectra of the CIB with respect to the CMB (both its purely Planckian part and its Compton distortion expected to be the dominant one at these frequencies) allow them to be separated very efficiently down to frequencies close to the peak of the CMB (150 GHz). Furthermore in this frequency regime the CIB dominates the galactic emission in the lowest cirrus regions by a factor 4. The cosmic background due to sources (CMB excluded) presents two maxima: one in the optical, one in the far-infrared, with roughly equal brightness (in I) and with a minimum around 5 µm. This minimum is created by the decrease of brightness of the stellar component with wavelength combined with the rising brightness of the dust, very small grains, and of the Active Galactic Nuclei (AGN) non thermal emission in the thermal infrared. The CIB is defined as the cosmic background at wavelengths longward of this minimum. An understanding of the nature and redshift distribution of the sources of the CIB, although relatively new, is an integral part of the understanding of the formation and evolution of galaxies.

The standard hierarchical model of structure formation has received strong observational support from observations of the large-scale distribution of galaxies, clusters, intergalactic clouds, combined with CMB anisotropies that constrain the initial large scale power spectrum within the concordance cosmological model framework. Scenarios for galaxy formation and evolution can be confronted with the very quickly rising set of observations of extragalactic sources at higher and higher redshifts. Nevertheless many critical questions remain open on the cooling of collapsed structures, angular momentum of galaxies, star formation efficiency, Initial Mass Function (IMF) of the stars formed, role of feedback mechanisms, the physics and role of merging and accretion in the construction of galaxies.

As ultraluminous infrared galaxies were found to be often associated with mergers or interacting galaxies, it can be expected that the sources of the CIB carry critical information about the history of merging (e.g., Sanders & Mirabel 1996; Genzel & Cesarsky 2000). Because about half of the energy from extragalactic sources is in the CIB, the determination of the source of this energy (starbursts or massive black hole accretion in dust enshrouded AGNs) should shed some light on how galaxies evolve.

Since the discovery of the CIB in the COBE data (e.g., Hauser & Dwek 2001), identifying the sources of the CIB, their redshift distribution, and nature progressed at increasing speed especially through multiwavelength analysis. This review attempts to give the broad-band observational picture for the identification of the sources of the CIB. The detailed analysis of individual infrared galaxies is outside the scope of this review. We then discuss implications, both well-established ones and tentative ones, as well as directions for future work. The spectroscopy aspects are covered in the review by Solomon & Vanden Bout (2005, in this volume). This is at present a fascinating but moving target! The coming decade will be a very rich one. New, very powerful long-wavelengths observation tools will bring many striking new results like the Spitzer observatory or the being built experiments like Herschel and ALMA. Throughout this review, the cosmology is fixed to = 0.7, _m = 0.3, H₀ = 100 h km s^-1 Mpc^-1 with h = 0.65.