Communication via quad/hexa-copters during disasters

Abstract

Purpose

The purpose is to derive the most effective place in the air for an aerial robot, viz., drone to use as an alternative communication system during disasters.

Design/methodology/approach

In this technology-driven era, various concepts are becoming the area of interest for multiple researchers. Drone technology is also one of them. The researchers, with interest in drones, are therefore trying to understand the various uses of employing drones in diverse applications which are mind-boggling, starting from civil applications (viz., an inspection of power lines, counting wildlife, delivering medical supplies to inaccessible regions, forest fire detection, and landslide measurement) to military applications (viz., real-time monitoring, surveillance, patrolling, and demining). However, one area where its usage is still to be exploited in many countries is using drones as a relay when communication lines are disrupted due to natural calamities. This will be particularly helpful in rescuing the affected people as the aerial node will enable them to communicate to the rescue team using mobiles/ordinary landline telephones even when regular communication towers are destroyed due to disastrous natural calamities, for example, tsunamis, earthquakes, and floods. Various algorithms, namely, water filling algorithm, advanced water filling algorithm, equal power distribution algorithm, and particle swarm optimization, were therefore studied and analyzed using simulation in addition to various path loss models to realize the desired place for an aerial robot, viz., drone in the air, which will eventually be used as an alternative communication system for badly hit ground users due to any disaster.

Findings

It was found that the effective combination of the water filling algorithm and particle swarm optimization algorithm may be done to place the drone in the air to increase the overall throughput of the affected ground users.

Originality/value

The research is original. None of the parts of this research paper has been published anywhere.

Citation

Vohra, D.S., Garg, P.K. and Ghosh, S. (2024), "Communication via quad/hexa-copters during disasters", International Journal of Intelligent Unmanned Systems, Vol. 12 No. 2, pp. 169-178. https://doi.org/10.1108/IJIUS-03-2023-0023

1. Introduction

Most of us understand drones as aerial base stations. They are generally assumed to capture photographs or create videos from various angles at various heights. However, very few thought of using drones as a relay for extending communication when regular lines of communication are destroyed during natural calamities, namely, tsunamis, floods, earthquakes, and landslides (Hildmann and Kovacs, 2019; Garg, 2021). In this work, we propose using the drone as a relay to provide communication between devices on the ground and a terrestrial base station situated far away from the reach of the mentioned ground devices (Shakoor et al., 2019). The aim is to provide a place for the drone in the air so that the affected ground communication devices (due to natural calamity) get again connected with this terrestrial network until the normal facilities are restored. It is otherwise very hard and next to impossible for the people stuck in such disaster-hit areas to connect to their near and dear ones. The situation keeps aggravating with every minute passing as the relatives from the connected places cannot connect, leading to chaos and confusion about the state of the people as the media channels keep exaggerating the situation by making it more sensationalized.

2. Past research

Many researchers have tried till now various deployment procedures based on reducing transmit power by increasing energy deficiency and the quality of service (QoS) (Shakhatreh et al., 2019; Zeng et al., 2019). Some researchers adhered to path loss models while analyzing the usage of drones in the form of a communication relay. At the same time, few researchers analyzed the same strategy for increasing drones' energy efficiency. Some researchers tried to deploy the drone to increase the reliability of services provided. The classification of the deployment of drones based on various functions is presented in Figure 1. The other functions (not illustrated in Figure 1) based on which the placement of drones in the air was also studied are generally based on communication stability when the user rides on the communication channel. The communication channel's stability depends upon the quality and bandwidth of the channel from the source end, namely, the drone (relay) to the destination end (Vohra et al., 2022). The stability also depends upon the termination equipment, namely, opti-mux, Synchronous Transfer Module (STM) 1, loop-mux.

3. Comparisons with existing work

A similar kind of research work was done by various researchers on the deployment of drones based on different functions, as illustrated in Figure 1. Some researchers only concentrated on the transmission power reduction, whereas others concentrated on the backhaul communication delay for homing on to the place for an aerial drone in the air. However, the remaining researchers tried to base their research on placing drones in the air totally on maximizing QoS in terms of triple play services, which means data, audio and video, or increasing the throughput of the drone transmission process (Tao et al., 2018). Post analyzing the comparisons, it can be concluded that no researcher has based his study on the placement of aerial drones on a water-filling algorithm, advanced filling algorithm or equal power distribution algorithm.

4. System model

4.1 Settings for the system

To make proper settings for the system, first, it is considered that there are a total of M wireless devices placed at some distance from the base communication station (Chen et al., 2017). The said M wireless devices, located far-flung, were therefore not communicating with the remote base station due to their restricted power emanating capability. Mathematically, let (X_u, Y_u and Z_u) are the coordinates where the drone will be placed (Alnagar et al., 2020). The ibid M wireless devices will start communicating with the remote base station via drone as a relay when the precise position for a drone is well analyzed and calculated. The same is pictorially shown in Figure 2. It is assumed that drones would use one of the multiple access techniques called frequency division multiple access. This will enable the drone in the air to send information at a signal-to-noise ratio (SNR), which will be greater than or equal to the least minimum SNR or the defined threshold value of SNR (SNR_th). Frequency division multiple access technologies will limit only one per-user subchannel for requisite communication (Yan et al., 2019). It will ultimately lead to no interference. It was also assumed that the highest communication power the aerial drone can transmit is p_max (see Figure 3).

4.2 Analysis of various path loss models

To analyze the path loss models effectively, these models were calculated first between the base remote communication station and drone and then between the drone (acting as a relay) and the ground users. The path loss model can be calculated between the drone and the ground user (Zhang et al., 2019; Ullah et al., 2019). Equation (1) describes the path loss model between the base station and an aerial relay node.

(1)LBS−UAV (d2D,θ) =10α⁡log(d2D)+A (θ – θo)exp(−(θ – θo)/ B) +ηo +N (0,aθ +σo)

where

• d_2D = √[(X_u − X_BS)² + (Y_u − Y_BS)²] is the two-dimensional distance between a base remote communication station and a drone,

• θ = angle of recession,

• α = path loss exemplar,

• A = scaler for excessive path loss,

• θ_o = offset for angle,

• B = scaler for angle,

• η_o = offset for excessive loss in the path,

• N (0, aθ + σ_o) is a random variable (Gaussian),

• in which “a” = shadowing of drone slope, σ_o = offset for shadowing of drone.

Equation (2) describes the path loss model between drone and ground users (i ∈ M), where i represents ground users with a maximum number of the same being M.

(2)LUAV−i (d3D,ϕ) =P (LOS). LLOS +P (NLOS). LNLOS

where

• d_3D = √ [(Xu – Xi)² + (Yu − Yi)² + Zu²] is the three-dimensional distance from drone to the ground user,

• P (LOS) = ensuring line of sight probability (LOS) at elevation ϕ,

• P (NLOS) = ensuring non-LOS probability. It is the same as (1-P (LOS)),

• L_LOS = average path loss for LOS paths,

• L_NLOS = path loss on average for NLOS.

4.3 Path loss model balance

A balance needs to be struck between the distances from the base remote communication station to the drone and from the drone to the ground user to leverage the path loss so that the best power allocation mechanism can be adopted (Sawalmeh et al., 2019; Kalantari et al., 2016). It was observed that whenever the distance between the drone to the ground user decreases, the path loss from the base station to the drone increases (Khuwaja et al., 2018). On the other hand, when the distance between the drone and the round user increases, path loss between the base station and the drone decreases. It is, therefore, of utmost importance to strike a balance between the above two-path losses, so that we can effectively place the drone at a place from where maximum ground users can benefit.

5. Formulation of mathematical problem

Let us consider triple-play communication services between the remotely placed communication base station and the ground user via drone as a relay. The capacity of the communication link between the remote communication base station and the drone is given in Equation (3).

(3)CBS−UAV=BBS. log2(1+SNRBS−UAV)

where

• B_BS = ground base station emanating bandwidth

• SNR_BS−UAV = drone's SNR

The capacity of the data

(4)CUAV−i=BUAV−i.log2(1+SNRUAV−i)

where

• B_UAV−i = drone's emanating bandwidth

• SNR_UAV−i = ground user's SNR

The maximum bandwidth of a ground user is estimated to be B_UAV/M, where B_UAV is the bandwidth of a drone, and M is the maximum number of wireless devices (Ullah et al., 2020; Sawalmeh et al., 2017).

Then, the required R_i data rate per wireless device are elucidated in Equation (5).

(5)pi=(2RiM/BUAV−1)N. LUAV−i

where various variables of equation (5) are defined in succeeding lines:

• L_UAV−i = losses on average between drone and ground user i

• N = power via noise

The exceptional placement of drones is ensured so that the maximum affected ground users stay in communication (Mozaffari et al., 2015; Al-Hourani et al., 2014a, b). The optimization problem is therefore elucidated below via equations (6) and (7).

(6)Maximize∑(i=1 to M)BUAV−i.log2(1+SNRUAV−i)

(xu,yu,zu,P)

(7)subject to∑(i=1 to M)BUAV−i.log2(1+SNRUAV−i)≤BBS.log2(1+SNRBS−UAV)

SNRUAV−i ≥SNRth,∀i ∈M,

∑(i=1 to M) pi ≤pmax,

pi ≥0,∀i ∈M,

xmin ≤Xu ≤xmax,ymin ≤Yu ≤ymax,zmin ≤Zu ≤zmax

The above equations specify various constraints starting from the capacity of the communication link to be less than or equal to the capacity of the communication link between the drone and the ground user. The next constraint, following the first constraint, ensures that the SNR of the ground user is always greater or equal to the threshold value, which is equal to SNR_th (Zhan et al., 2011). The third constraint set is formed so that the total power taken by the drone, in no way, should be greater than p_max, which is the maximum emanating level. The fourth constraint is set to ensure that the power consumed by drones is always greater than or equal to zero. The residual restraints represent the values for the coordinate that the drone would be placed into. The scenario description is shown pictorially in Figure 4.

6. Methodology

First, the drone placement was done with the help of the particle swarm optimization (PSO) algorithm. The velocity and placement of the particle are revised in each iteration. The time complexity of the implemented algorithm was fully dependent on the number of iterations. In this work, an objective function was considered where the drone is static at a particular height. Post assessing the same, the throughput of the affected users who were affected by the disaster was analyzed by the equal power distribution algorithm, water filling algorithm and modified water filling algorithm for achieving the maximum throughput by affected users. It was found that the water filling algorithm was the most effective even when the distance between ground users and the drone was more than 3000 m. However, up to 4000 m, the water filling algorithm behaved the best. Figure 5 depicts that the water filling algorithm is the best, as it provides efficacy and the same effect irrespective of the distance. The water filling algorithm uses an interchannel interference method so that the best channel is always available to the user, even at greater distances. The channels are orthogonal in nature so that there is no interference at any point in time. The complexity of the water filling algorithm is also easier to implement in terms of simulation comparison of advanced water filling algorithm and equal power distribution algorithm.

7. Simulation parameters

The simulation parameters set for concluding are enunciated in Table 1.

The algorithmic parameters set for concluding are enunciated in Table 2.

8. Water filling algorithm

Finally, the water filling algorithm is implemented in detail to analyze the optimum power allocation of drones between ground users, post comparing the water filling algorithm with equal power distribution and modified water algorithms (as stated in section 6.0). To understand the same, Figure 6 has been drawn to show the user distribution inside the affected area R. In the painted manipulated scenario, it was assumed that the total transmission bandwidth is X MHz. For a user to draw the power for adequate transmission and reception, the figure of X/1000 arrived, that is, if 50 MHz total bandwidth is available and a user gets a bandwidth of 50 Kbps. It was assumed that the ground user would be able to communicate well (Palomar and Fonollosa, 2005). The other equal power allocation algorithms were also analyzed to draw a comparison with other similar power allocation algorithms (Al-Hourani and Gomez, 2017). It was observed mathematically that the water filling algorithm outperforms the ergodic capacity for the same communication channels (He et al., 2020). The concept was to arrive at a threshold value the drone should transmit (Perera et al., 2020). Moreover, the strategy was to transmit power more in good channels between the drone in the air and the ground user. It doesn't mean transferring zero transmission for bad channels. A balance needs to be struck so that the optimum power is transferred in good channels. However, it was also considered that the bad channels could not be neglected either (Kumar et al., 2018). The case was therefore analyzed where the transferring power at a value below the threshold value was experimented with and then analyzed. However, it did not make sense to transmit negative power and therefore neglected the fundamental concept of the water filling algorithm. The water filling algorithm also ensures maximum coverage for users compared to other power allocation algorithms, even at distances beyond 3000 m.

9. Results and discussion

Post-simulation, it can be concluded that the water filling algorithm performed the best at distances of more than 4000 m. Moreover, the water filling algorithm performed inferior to 4000 m compared to the advanced and equal power distribution algorithms. However, practically the disaster-hit-prone area would be covering an area of more than 5,000–10000 m at least, so practically, the decision to arrive based on an advanced water filling algorithm and equal power distribution algorithm will serve no purpose. Hence, the situation in disaster-prone hit areas was understood practically before reaching a well-informed conclusion in this research work. Theoretically, otherwise, it would have incorrectly concluded that the water-filling algorithm is performing the worst in placing drones in the air for the relay purpose in disaster-hit areas.

10. Conclusion

A drone planned as a relay media in the air is one of the possible solutions to connect to the users where presently no connectivity exists or is disrupted during disasters. Before arriving at the notion, a detailed analysis was done using path loss models, the basic building blocks of any communication. Therefore, a cellular-to-drone path loss model was analyzed in detail for the backhaul connectivity between the ground base station and the drone. In contrast, the path loss model between the drone and ground users was examined in totality for the downlink connection between the drone and ground users. Subsequently, the balance between both the path models was struck. Therefore, finding the most effective place for drones was formulated as a convex-optimized mathematical problem with various constraints. The objective set for the optimization problem was to increase the throughput of the ground users. It was observed that the water filling power allocation algorithm best suits the situation for effectively homing onto the most effective place for the drone poststationing drone with the PSO algorithm. As a sequel of this approach, the reverse, that is, the uplink communication link from ground users to drones may be analyzed in detail.

10.1 Future works

As a continuation of this work, the present research can be further studied based on the uplink communication link from ground users to the drones in the air. Further, a stochastic average analysis will be carried out based on the present research work and uplink communication link from ground users to the drone side. It will lead to a more informed analysis. It will lead to a better decision about positioning drones in the air for relaying communication in disaster-hit areas.

Figures

Figure 1

Deployment classification of the drone as a relay

Figure 2

Path loss model trade-off

Figure 3

System model

Figure 4

Scenario description of remotely placed base station, drone, and ground users

Figure 5

Comparison between modified water filling, water filling and equal power distribution algorithms

Figure 6

Distribution of users inside the affected area R

Table 1

Simulation parameters

Dimensions of area R	(X1, X2, Y1, Y2)	(0,1000,0,1000)
Maximum number of affected users due to calamity	M	100–1,000
Transmit Max of drone	P(drone)	30 mW
Transmit Max of affected users	P(user)	46 mW
Bandwidth of drone	BW (drone)	50 MHz
Bandwidth of user	BW (user)	75–100 MHz
Location of remote communication base station	Loc (remote_base_stn)	(7,000, 500)

Source(s): Prepared by Authors

Table 2

Algorithmic parameters

Frequency of carrier	Fc	2 GHz
PSO size	PSO (population)	50
Number of iterations	Iteration (PSO)	50
Other random parameters	(a, b)	(9.2, 0.8)
Power of noise	Noise(power)	−120 mW
Threshold Signal to Noise	SNth	35 dB

Source(s): Prepared by Authors

Competing interests: There are no financial and non-financial competing interests with any person, party, office/ department/ institute/ organization.

Availability of data and materials: Data supporting the findings are part of the manuscript only.

Funding: No funding from any source of any nature was obtained.

Authors’ contributions: All authors have contributed to ensuring that sound conclusions may be reached.