Secure Communications Over Cell-Free Massive MIMO Networks With Hardware Impairments

Jamming-Robust Uplink Transmission for Spatially Correlated Massive MIMO Systems

“…We cannot generally show that the objective function in L is concave, however the optimal values p o t and p o d can be found numerically. This should be taken into consideration that the relaxation methods, such as those proposed by [17], [18] for the uncorrelated fading channels, are impractical in our setup with the correlated fading channels. Moreover, we can analytically obtain useful insights into p o t and p o d in one interesting scenario that is P w ≫ P ≥ 1.…”

“…We plot the SE of the MS approach for all 3 users versus the jamming powers q t = q d in Fig. 4 for ideal hardware and two cases that are perfect knowledge about S (17) and (18) nearly behave the actual SEs for all 3 users. Accordingly, it is expected that the achievable SEs can be realized without the perfect knowledge of the jamming channel statistics.…”

“…Wang et al in [15] assign the beamforming coefficients such that the approximate secrecy rate is maximized, while in [16], the asymptotic secrecy rate is derived assuming matched filter beamforming and artificial noise generation at the BS. Optimal power control for preventing eavesdropping in the downlink transmission of the cell-free massive MIMO systems is proposed in [17] for ideal hardware and in [18] for non-ideal hardware. Previous works have either considered security problem in the presence of active eavesdropper [14]− [18] or jamming suppression problem for only uncorrelated Rayleigh fading channels [10]− [13].…”

“…Optimal power control for preventing eavesdropping in the downlink transmission of the cell-free massive MIMO systems is proposed in [17] for ideal hardware and in [18] for non-ideal hardware. Previous works have either considered security problem in the presence of active eavesdropper [14]− [18] or jamming suppression problem for only uncorrelated Rayleigh fading channels [10]− [13]. In general, the jammer's strategy depends on its objective and its available information.…”

“…In general, the jammer's strategy depends on its objective and its available information. Accordingly, in [5], [6], [11], [12], [15], the authors suppose that the pilot jamming attack is a combination of several pilot sequences of the pilot set, while in [4], [7]− [10], [14], [16]− [18], the jammer selects a target user's pilot sequence as the jamming pilot.…”