LAUSR

The emerging collaborative authentication schemes are capable of outperforming the conventional isolated methods as a benefit of their multi-dimensional data/information gleaned, but they face new challenges in the sixth generation (6G) wireless networks owing to their increased overhead, limited flexibility and autonomy. Moreover, they may also be vulnerable to the privacy leakage of individual entities. These challenges are mainly due to the complex heterogeneous network architecture, owing to the distributed nature of the devices and information involved as well as the diverse security requirements of the 6G-aided vertical systems. As a remedy, we introduce autonomous collaborative authentication for achieving security enhancement through the situation-aware cooperation of different security mechanisms, of heterogeneous security information/context, and of heterogeneous devices and networks. For this purpose, a federated learning-based collaborative authentication scheme capable of privacy-preservation is developed, where cooperative peers observe and locally analyze heterogeneous information of the authenticating device, and afterwards update their authentication models locally. By sharing their authentication models rather than directly sharing the observed authentication information, privacy preservation can be achieved based on the proposed scheme. Moreover, given the time-varying heterogeneous network environment and the wide range of quality-of-service (QoS) requirements, the membership of the group collaborating in support of distributed authentication is updated based on the situation-dependent conditions. To further reduce the communication overhead, a locally collaborative learning process is further developed, where both the updated parameters and observed authentication information are stored and processed locally at the cooperative peers. Finally, a smart contract is designed for achieving collaborative security combined with privacy preservation and for providing accountable services.