A bistatic measurement technique is implemented with an authentication approach based on tag-backscattered electric field (E-field) measurements at different orientation angles for unitary classification in the millimeter-wave (mmWave) band. The idea is based on the augmentation of information related to the aspect-independent tag-backscattered signals according to different angle measurements. Geometric uncertainties inherent to the manufacturing process are transcribed in minor variations observed in the tag electromagnetic response and exploited from the measurement at three different angles to authenticate the tag. This information in relation to the angle increases the authentication level. A set of sixteen E-shape chipless tags fabricated to operate at millimeter-wave frequency are analyzed. To better exploit a large amount of data collected with this approach, a Machine Learning (ML) classification is evaluated. The probability of error (PE) achieved with the method is around $0.05\%$. This PE is the lowest obtained considering chipless RFID tags.