| Autor: |
Weijia Lei, Yang Zhou, Xiuzhen Lin |
| Jazyk: |
angličtina |
| Rok vydání: |
2021 |
| Předmět: |
|
| Zdroj: |
Digital Communications and Networks, Vol 7, Iss 3, Pp 352-361 (2021) |
| Druh dokumentu: |
article |
| ISSN: |
2352-8648 |
| DOI: |
10.1016/j.dcan.2020.07.004 |
| Popis: |
We discuss the physical layer security scheme in the Full-Duplex (FD) MIMO point-to-point two-way communication system with residual self-interference, in which legitimate nodes send confidential information and null space Artificial Noise (AN) while receiving information. Because the Channel State Information (CSI) of the eavesdropper is unavailable, we optimize the covariance matrices of the information signal as well as the allocation of the antenna for transmitting and receiving to minimize the signal power consumption under the target rate constraint. As a result, the power of AN is maximized within the limit of total power, so the interception capability of the eavesdropper is suppressed as much as possible. Since self-interference cannot be completely eliminated, the optimization process of one legitimate node depends on the optimization result of the other. By substituting self-interference power in the secrecy rate formula with its average value, the joint optimization process at the two nodes is transformed into two separate and solvable optimization processes. Then, the Water-Filling Algorithm (WFA) and bisection algorithm are used to get the optimal covariance matrices of the signal. Furthermore, we derive the theoretical lower bound of ergodic achievable secrecy rate under rayleigh channels to evaluate the performance of the scheme. The simulation results show that the theoretical derivation is correct, and the actual achievable rate is very close to the target rate, which means that the approximation in the optimization is feasible. The results also show that secrecy transmission can be realized because a considerable secrecy rate can be achieved. |
| Databáze: |
Directory of Open Access Journals |
| Externí odkaz: |
|
