Cell-free massive MIMO with multiple active eavesdroppers

This paper investigates the secrecy performance of cell-free massive MIMO (CF-mMIMO) systems in the presence of active spoofing attacks by multiple eavesdroppers (Eves). Each Eve conducts a spoofing attack on a different legitimate user during the uplink training phase, aiming to intercept the information intended for that user during the downlink transmission phase. To counter these attacks, we propose a joint access point (AP) selection and power optimization strategy to enhance the security performance of the CF-mMIMO system. Specifically, we formulate an optimization problem that seeks to maximize the sum-spectral efficiency (SE) of the legitimate users while ensuring a positive secrecy spectral efficiency (SSE) for all the attacked users. A sub-optimal solution to this mixed-integer non-convex problem is obtained using an efficient, low-complexity accelerated projected gradient (APG)-based algorithm. Moreover, for system design purposes, we introduce two simple and efficient methods: i) detecting the presence of multiple active Eves within the system and identifying which users are under attack, and ii) estimating the large-scale fading coefficients between the APs and the Eves. Our findings show that the proposed approach achieves a median sum-SE performance that is 62% better than that of equal power allocation without AP selection scheme. Furthermore, the results demonstrate that the proposed strategy significantly improves the sum-SE while ensuring a positive secrecy rate, thereby safeguarding the confidentiality of all transmitted information signals to all users, even in the presence of a relatively large number of Eves.