Inflationary models were born out of grand unified theories (GUTs) where a mechanism involving symmetry breaking was used to drive it. Also out of these same theories another group of models have merged which are radially different to inflation. These are the models of topological defects, which are produced between regions which ``freeze out'' with different phases during symmetry breaking. There are many types of defects: monopoles, strings, sheets and textures. The one which have generated the most attention is that of strings, which are predicted by Kaiser and Stebbins[5] to produce sharp line-like discontinuities due to the effect of the changing gravitational potential of a moving string on passing CMB photons. A similar effect would happen to the primordial matter fields deflecting particles into wakes which collapse and form galaxies. More recently Couslson et al.[4] have made numerical simulations of the effects of monopoles and textures and finds these produce distinctive hot and cold spots rather than lines. The power spectrum of these models gives a scale invariant spectrum similar to that of inflation up to where the first Doppler peak in expected, but there can not be Doppler peaks as there are no acoustic oscillations in the early universe to generate them.
Another signature of defects which is not apparent in the power spectrum is non-Gaussianity. Inflation predicts a Gaussian random field due to the superposition of a near infinite number of differently stretched quantum fluctuation modes, but defects cause isolated sharp features where their gravitational influences are strongest. The power spectrum just measures the average spectral density over angular frequencies, so the extreme but infrequent hot and cold spots have relatively little impact. The only why to check for these is to make a high signal to noise map and would make very interesting examples of COSMOSOMAS.