This paper investigates the performance analysis of cognitive-inspired wireless-powered nonorthogonal multiple access (NOMA) systems, where two collaborative schemes are jointly proposed to enhance the quality-of-signal at the primary cell-edge (CE) user in dead zones. In particular, the first scheme, namely internal collaborative schemes, is developed based on the availability of primary cell-near users (CU), while the rest, called external collaborative schemes, comes from the help of a secondary source (SS). For the second scheme, SS' operation is considered with two relaying-encoding NOMA protocols, namely decode-and-forward (DF) and amplifier-and-forward (AF). In order to enable CE's communications in short execution times, the power-splitting mechanism is considered at both CU and SS. To capture the system's performance trend accurately, closed-form approximate expressions for the outage probability of users are derived. Moreover, we also carry out the respective asymptotic expression at high transmit power regimes to show some useful insight, such as diversity order and coding gain. Finally, extensive Monte-Carlo simulations are presented to collaborate the accuracy of the theoretical analysis and numerical results show that the proposed collaborative scheme provides CE's performance enhancement better than using only one collaborative scheme.INDEX TERMS Amplify-and-forward relaying protocol, cognitive radio, energy harvesting, non-identical fading channel, non-orthogonal multiple access (NOMA), overlay spectrum sharing, simultaneous wireless information and power transfer (SWIPT).