Investigation of a HAP-UAV Collaboration Scheme for Throughput Maximization via Joint User Association and 3D UAV Placement

. 2023 Jul 2;23(13):6095. doi: 10.3390/s23136095

Algorithm 4 Genetic-Algorithm-Based Allocation.

1:
Find $| M |$ clusters and centroid
2:
Init-UAV-placement( $x_{m}, y_{m}, z_{m}$ ) ← centroid
3:
LB ← [ $x_{m} - r n g_{g} y_{m} - r n g_{g} h_{m i n}$ ]
4:
UB ← [ $x_{m} + r n g_{g} y_{m} + r n g_{g} h_{m a x}$ ]
5:
Genetic algorithm’s parameters ← default
6:
Find UAVs’ placement using the genetic algorithm with the objective to maximize the total throughput, which is the output of the following function.
7:
8:
Function Throughput-calculation
9:
Throughput = zeros( $| M | + 1, | U |$ )
10:
for $m = 1 t o | M |$ do
11:
Throughput(m,:) = NLoS throughput Formula (8), using positions given by one generation of the genetic algorithm.
12:
end for
13:
Throughput( $| M |$ + 1, :) = LoS throughput Formula (5), using positions given by one generation of the genetic algorithm
14:
A = zeros( $| M |$ + 1, $U$ ), Throughput-temp = Throughput
15:
while sum(A) $< | U |$ do
16:
Find max(Throughput-temp) & its index-x and index-y
17:
A(index-x,index-y) = 1
18:
Delete Throughput-temp(:,index-y)
19:
if index-x ≤ $| M | &$ sum(A(index-x,:)) = $N_{m a x}$ then
20:
Delete Throughput-temp(index-x, :)
21:
end if
22:
end while
23:
Total-throughput = sum(Throughput ⊙ A)
24:
return Total-throughput
25:
Endfunction