Feng Shu,Zaoyu Lu,Jinyong Lin,Linlin Sun,Xiaobo Zhou,Tingting Liu,Shuo Zhang,Wenlong Cai,Jinhui Lu,Jin Wang
Abstract:In unmanned aerial vehicle (UAV) networks, directional modulation (DM) is adopted to improve the secrecy rate (SR) performance. Alice, a ground base station, behaves as a control center, and Bob is a UAV of flying along a linear flight trajectory who optimizes its SR performance by dynamically adjusting its beamforming vectors and power allocation (PA) strategy. Per fixed time interval during the Bob's flight process, the transmit beamforming vectors for useful messages and AN projection are given by the rule of maximizing signal-to-leakage-and-noise ratio (Max-SLNR) and maximizing AN-and-leakage-to-noise ratio (ANLNR), and the optimal PA strategy is based on maximizing SR (Max-SR). More importantly, an alternating iterative structure (AIS) between beamforming and PA is proposed to further improve the SR performance. Simulation results show that the proposed AIS converges rapidly, and can achieve substantial SR gains over Max-SLNR plus Max-ANLNR with fixed PA such as PA factor $\beta=$ 0.5, and 0.9. In particular, in the case of small-scale antenna array, the SR performance gain achieved by the proposed AIS is more attractive. Additionally, as the number of antennas tends to be large-scale, the average SR performance of the proposed AIS approaches an SR ceil.
What problem does this paper attempt to address?
This paper attempts to solve the problem of how to improve the secrecy rate (SR) through beamforming and power allocation strategies in an unmanned aerial vehicle (UAV) - assisted direction - modulation network. Specifically, the paper focuses on the secure communication problem between the ground base station (Alice) and the UAV (Bob) moving along a straight - line flight trajectory.
### Main problems
1. **Improving the secrecy rate**: Due to the broadcast characteristics of wireless signals, achieving secure transmission in UAV networks is a challenge. To this end, the paper proposes an alternating iterative structure (AIS) to optimize the secrecy rate by dynamically adjusting the beamforming vector and power allocation strategy.
2. **Joint optimization of beamforming and power allocation**: Traditional research usually optimizes the beamforming scheme and power allocation strategy independently, while this paper attempts to establish the relationship between the two to achieve information transfer and collaborative optimization.
### Solutions
- **Design of beamforming and artificial noise projection vectors**: To simplify the complex joint optimization problem, the paper first designs the beamforming vector \( \mathbf{v}_{b,n} \) and the artificial noise (AN) projection vector \( \mathbf{v}_{AN,n} \) according to the leakage criterion.
- **Optimal power allocation strategy**: Given the beamforming vector and the AN projection vector, the optimal power allocation factor \( \beta_n \) is determined by maximizing the secrecy rate (Max - SR).
- **Alternating iterative structure (AIS)**: The paper proposes an alternating iterative structure, that is, in each iteration, first calculate the beamforming vector and the AN projection vector according to the current power allocation factor, and then recalculate the power allocation factor according to the new beamforming vector and the AN projection vector until the termination condition is met.
### Models and formulas
1. **Received signal model**:
- The signal received by Bob is:
\[
y_{b,n} = \sqrt{g_{ab} \beta_n P_s} \mathbf{h}^H(\theta_{b,n}) \mathbf{v}_{b,n} x_n + \sqrt{g_{ab} (1 - \beta_n) P_s} \mathbf{h}^H(\theta_{b,n}) \mathbf{v}_{AN,n} z_n + w_{b,n}
\]
- The signal received by Eve is:
\[
y_{e,n} = \sqrt{g_{ae} \beta_n P_s} \mathbf{h}^H(\theta_{e,n}) \mathbf{v}_{b,n} x_n + \sqrt{g_{ae} (1 - \beta_n) P_s} \mathbf{h}^H(\theta_{e,n}) \mathbf{v}_{AN,n} z_n + w_{e,n}
\]
2. **Secrecy rate calculation**:
- The achievable rate \( R_{b,n} \) of Bob is:
\[
R_{b,n} = \log_2 \left( 1 + \frac{g_{ab} \beta_n P_s \mathbf{h}^H(\theta_{b,n}) \mathbf{v}_{b,n} \mathbf{v}_{b,n}^H \mathbf{h}(\theta_{b,n})}{g_{ab} (1 - \beta_n) P_s \mathbf{h}^H(\theta_{b,n}) \mathbf{v}_{AN,n} \mathbf{v}_{AN,n}^H \mathbf{h}(\theta_{b,n}) + \sigma_b^2} \right)
\]
- The achievable rate \( R_{e,n} \) of Eve is:
\[
R_{e,n} = \log_2 \left( 1 + \frac{g_{ae}