Bio-Inspired Algorithms for Efficient Clustering and Routing in Flying Ad Hoc Networks

. 2024 Dec 26;25(1):72. doi: 10.3390/s25010072

Algorithm 4: Congestion-aware fault-tolerant routing.

Input : Positions of UAVs P_{i} = (x_{i}, y_{i}, z_{i})

, velocity of UAVs

({V x}_{i}, {V y}_{i}, {V z}_{i}

), residual energy level of UAVs E (i)

, congestion data from ferry nodes, queue length Q L

, packet rate P R

, and neighbor table N T

Output: Primary path and secondary path.

# Step 1: Initialize UAV parameters

1 : for each UAV i

do:

2 : P_{i} = (x_{i}, y_{i}, z_{i})

V_{i} = ({V x}_{i}, {V y}_{i}, {V z}_{i})

E (i) = E_{i}

, {Q L}_{i} = 0

, {P R}_{i} = 0

, and {N T}_{i} = {}

3: end for
# Step 2: Periodically send Hello Messages and collect neighbor data

4 : for each UAV i

do:

5: Broadcast (Hello Message)

6 : end for

7 : for each UAV i

do:

8 : Receive Hello Message from neighbor j

9 : Update {N T}_{i} = {P_{j}, V_{j}, E_{j}, {Q L}_{j}, {P R}_{j}}

10 : end for # Step 3 : Ferry nodes collect global information (density and congestion) 11 : for each ferry node do : 12 : Collect positions of all UAVs within communication range

13 : Calculate the r e g i o n a l d e n s i t y = \frac{N u m b e r o f U A V i n A r e a}{A r e a s i z e}

14 : Share the density and congetion level with all UAVs 15 : end for # Step 4 : Eligibility phase based on energy threshold

16 : for each UAV i do :

17 : Eligible nodes = [] 18 : for each neighbor j of UAV i do :

# Check if neighbor j

has sufficient residual energy

19 : if E_{j} \geq E_{t h then}

20 : Eligible nodes append (j

) / / Add j

to list of eligible nodes

21 : end if

22 : end for

23 : end for

# Step 5 : Fitness evaluation for eligible nodes (primary path)

24 : for each UAV i

do :

25 : Best fitness value B F V = - 1

26 : Selected forwarder = None

27 : for each eligible neighbor j

in eligible nodes do

# Predict future position of UAV j

based on mobility

28 : Predict the future position using (18)

# Calculate distance from future position to destination

29 : Calculate the distance from future position to destination using (19)

# Calculate directionality score (if moving toward destination)

30 : Calculate the direction {D i r}_{j} = (V_{j}, (d e s t i n a t i o n - P_{j})

31 : if {D i r}_{j} > 0

then

32 : Calculate the mobility score and load for neighbor j

using (20) and (21)

33 : Calculate the fitness value using (22)

34 : if {F V}_{j} > B F V

then
35

B F V = {F V}_{j}

and select

forwarder node j

36: end if
37: end if
38: end for
39: end for
# Step 6: Congestion monitoring by ferry nodes for secondary path activation
40:

for each ferry node do :

41 : Collect Q L, P L, U D

and measure the congestion level by {C L}_{n e t} = f (Q L, P L, U D)

42: for each UAV

i

in region do

43 : Send C L

to UAV

44 : end for

45 : end for

# Step 7 : Decision logic for fault - tolerant \sec ondary path activation

46 : for each UAV i

do :

47 : if {C L}_{n e t} > {C L}_{t h}

then
48:

Activate A R P = T r u e

49 : else Activate A R P = F a l s e

50 : end if

51 : end for

# Step 8 : Sec ondary path selection for congestion avoidance (if activated)

52 : if activate A R P

then

53 : for each eligible neighbour k

of UAV i

not on the primary path do

54 : Ensure geographic separation from primary path using (24)

55 : if direction seperation k > A_{t h}

then

56 : Calculate the load for \sec ondary path candidate k

57 : Calculate the fitness function for \sec ondary path candidate k

using (25)

58 : if {F F}_{k} > B F V

then
59:

B F V = {F F}_{k}

60: end if
61: end if
62: end for
63: end if
# Step 9: Forward data via selected paths
64:

for each UAV i

do :

65 : if activate \sec ondary path A R P

then
66: Forward packet (selected forwarder) // Primary Path
67: Forward packet (selected secondary forwarder) // Secondary Path
68: else forward packet (selected forwarder) // Only Primary Path
69: end if
70: end for