Algorithm 2 Mobility Control method for multi-agent-inter communication

Require:${\mathit{p}}_{t_0}^{\ast }={\mathit{p}}_0$, ${\psi }_{t_0}$.

1: for $t_k=t_1,\dots ,t_K$ do

2:   Predict user positions $\mathit{P}=\left\{{\mathit{p}}_i\right\}$ at $t_{k+1}$ using Kalman Filter.

3:   Estimate uplink channel parameters with the approximately model defined in Equation (33) using the LSE algorithm, based on observations during time $[t_{k-1},t_k)$.

4:   ${\widehat{\mathit{p}}}_{k+1}^{\ast }={\mathit{p}}_k+G\ast \mathrm{dir}(\nabla J)$.

5:   UAV flies towards ${\widehat{\mathit{p}}}_{k+1}^{\ast }$ during time interval $[t_k,t_{k+1}]$ with $\dot{\psi }=-K_l\langle \eta -{\eta }^{\mathrm{d}}\rangle +\frac{{\dot{\eta }}^{\mathrm{d}}}{k_{\eta }(\psi )}$.

6: end for