Some limit behaviors for the LS estimator in simple linear EV regression models

In the present paper, we study the simple linear errors in variables (EV) model: [eta]i=[theta]+[beta]xi+[epsilon]i,[xi]i=xi+[delta]i, with i.i.d. errors . The consistency and asymptotic normality for the LS estimators and of the unknown parameters [beta],[theta] are obtained, which weaken some known conditions and improve some known results. Finally, the large deviation principle for and are given under the assumptions that ([epsilon]i,[delta]i) possess normal distributions.