| One of the major interests of current cosmology is to determine whether the present large-scale structure (LSS) has been formed by cosmic defects, such as strings and textures, or by quantum fluctuations generated during inflation. In this thesis, I study the roles of defects on LSS formation with or without a previous inflation period.; This thesis has two main topics. One is to investigate whether defect theories alone (without inflation) can reproduce the structure of the present universe. I re-examine matter power spectrum due to cosmic defects, using updated LSS observation data. Results show that in an {dollar}Omegasb0{dollar} = 1 universe, the cosmic string model is almost certainly ruled out. This model is not likely to work in an open universe either, although further study is required to rule it out conclusively.; The texture model in an {dollar}Omegasb0=1{dollar} universe is more difficult to rule out than the string model. Indeed I conclude that it is premature to reject the model because of its theoretical and observational uncertainties, although there are several negative indications. I also find that the texture model is not likely to be saved by making the universe open (without a cosmological constant), although the model may work in a flat cosmological constant dominated universe.; The second topic of this thesis deals with defect models with an inflationary period before defect formation (i.e., pre-inflation). I investigate particle physical models in which such mixture of defect and inflation theories are realized "naturally". Then, power spectra of these mixed models are compared with LSS observation data. I show that some inflation models, which are inconsistent with these data otherwise, can be saved by this mixture. |
