Page 55 - ITU Journal Future and evolving technologies Volume 3 (2022), Issue 2 – Towards vehicular networks in the 6G era

ITU Journal on Future and Evolving Technologies, Volume 3 (2022), Issue 2

interferers (denoted by “MI1”, “MI2”, “MI3”) moving Next, we apply the GBMU algorithm to the vehicles
at the same velocity, the location of the transmitter simulated in our Stage-1 network level simulator,
(denoted by “UE0” in the figures) is updated where there are 21 UEs in the network and UE0 is
iteratively according to the algorithm with step size the vehicle whose packet reception rates before and
= 1. UE0 finally reaches a point in the network and after the moving updates are recorded to visualize
the moving distance per iteration diminishes as the the convergence. Fig. 7 shows the iterative position
iteration proceeds. In order to verify the updates with the step size = 1 for one the
convergence and the effectiveness of the algorithm, snapshots where UE0 is originally at a non-centered
after each movement of UE0 in iteration k, the new position with respect to all other vehicles. With the
position will be used to calculate the path loss GBMU algorithm, UE0 can find itself a better
,
and so on the ( ) of each link, and finally the position to stay in order to maximize its utility,
,
normalized utility gain at iteration k could be which takes the success packet reception
computed as: probability for all its potential receivers and
fairness among them into account. The GBMU
( )− (0)
( ) = , (17) algorithm takes UE0 finally to reach a better

| (0)|
position in this vehicle network and stay there as
In this function, ( ) is the new utility gain and the algorithm smoothly converges with a utility

(0) is the initial utility gain. The results obtained increase by 45.6% at its convergence shown in

are shown in Fig. 6. It can be observed that after 500 Fig. 8. Please note that the final position can be
times position updates, the algorithm shows found by the node UE0 via in-node computation to
convergence and achieves a utility gain of 42.2% in go over the iterations, instead of really moving itself
this case. in the network before the convergent position is
obtained.

400
500
MI2
350 450
UE13
400
300
350
UE0(start) UE12
250 UE4 UE14
UE2 300 UE0(start) UE1
meters 200 Meters 250 UE10 UE9
UE7
UE3 200 UE0(end)
150
UE0(end)
UE15 UE5 UE3 UE6 UE16
MI3 UE1 150
100 UE11 UE2
100
50 MI1
50
UE8
0 0
0 50 100 150 200 250 300 350 400 0 50 100 150 200 250 300 350 400 450 500
meters Meters
Fig. 5 – Moving update model Fig. 7 – Moving update model with all vehicles
45 0.5
40 0.45
35 30 0.35
0.4
Normalized utility gain (%) 25 20 Normalized utility gain (%) 0.25
0.3

0.2
10 15 0.15
5 0.1
0 0.05
0 500 1000 1500 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000
Iteration Iteration
Fig. 6 – Gain convergence Fig. 8 – Gain convergence with all vehicles

© International Telecommunication Union, 2022 43