Abstract
Frequency division duplex (FDD) massive multiple-input multiple-output (MIMO) systems introduce a large overhead in downlink channel estimation in contrast to the time division duplex (TDD) mode. This overhead results in a considerable spectral efficiency (SE) gap between the FDD and TDD modes. In this paper, we consider the performance of the TDD and FDD massive MIMO systems with a spatially correlated channel in the presence of jamming in the network. We show how a smart jammer can effectively design its attack signal to degrade the network performance in terms of the downlink SE. Since the jammer can obtain different information about the channels in the TDD and FDD modes, two distinct jamming strategies are proposed for each mode. In the numerical results, the performance of both the TDD and FDD modes under the optimized jamming designs are evaluated and compared. Our results show that despite more attention to the TDD mode in current massive MIMO systems, it is more vulnerable to smart jamming attacks compared to the FDD mode which results in a smaller SE-gap between the modes. Furthermore, a countermeasure technique is proposed to combat this jamming attack in both the TDD and FDD modes by estimating the jamming power, grouping users, and allocating power among them. The numerical results demonstrate the effectiveness of the proposed countermeasure techniques in both the TDD and FDD modes.
Highlights
Massive multiple input multiple output (MIMO) technique is one of the main solutions to meet the huge capacity demands in 5G networks
We address the problem of physical layer security in massive MIMO networks in both the time division duplex (TDD) and Frequency division duplex (FDD) modes
We show that the TDD mode is more vulnerable against a smart jamming attack compared to the FDD mode, because the smart jammer can obtain more information about the channels in the TDD mode
Summary
ASHKAN SHEIKHI 1, (Member, IEEE), S. MOHAMMAD RAZAVIZADEH 1, (Senior Member, IEEE), AND INKYU LEE 2, (Fellow, IEEE) This work was supported by the National Research Foundation through the Ministry of Science, ICT, and Future Planning (MSIP), Korean Government, under Grant 2017R1A2B3012316.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have