• Providing a lightweight RFID-based authentication approach for IoT application to address the computation and hardware limitations. • Applying nonlinear key-stream generators to eliminate the linearity property of LFSR-based generators. • Evaluating the efficiency of the proposed approach based on security analysis and statistical analysis. • Assessment of the effect of generator polynomial degree and the proposed combinational generators on the BER using the Cooja tool. • Assessment of overhead of the proposed approach in terms of computational cost, storage spaces, and communication costs. In this paper, a new lightweight RFID-based authentication approach is proposed that uses stream ciphering techniques to preserve privacy between legitimate components. The proposed approach uses various key-stream generators such as And2LFSR, Geffe, and majority-based to eliminate the inherent linearity of LFSRs and construct a non-linear Pseudo-Random Number Generators (PRNG). The informal security analysis indicates that the proposed approach provides forward security and resists eavesdropping, tag tracking, replay, cloning, and DoS attacks. Moreover, simulations conducted through Cooja tool illustrate the effect of generator polynomial degree and proposed combinational generators on BER. The BER of communications has been evaluated in different quantities of (E b / N 0), ranged between 1 and 10, using the proposed generators along with different error control techniques. In addition, evaluations indicate that the proposed approach outperforms in terms of computational cost, storage spaces, and communication cost. Furthermore, statistical analysis proves resistance of the proposed approach to guessing attacks. [Display omitted] [ABSTRACT FROM AUTHOR]