Energy-Efficient User Pairing and Power Allocation for Granted Uplink-NOMA in UAV Communication Systems
محورهای موضوعی : Wireless Network
Seyed Hadi Mostafavi-Amjad
1
(Faculty of Electrical Engineering, Urmia University of Technology, Urmia, Iran )
Vahid Solouk
2
(Department of IT and Computer Engineering, Urmia University of Technology, Urmia, Iran)
Hashem Kalbkhani
3
(Faculty of Electrical Engineering, Urmia University of Technology, Urmia, Iran )
کلید واژه: Energy Efficiency, NOMA, Power Allocation, Unmanned Aerial Vehicle (UAV), Uplink, Users Pairing,
چکیده مقاله :
With the rapid deployment of users and increasing demands for mobile data, communication networks with high capacity are needed more than ever. Furthermore, there are several challenges, such as providing efficient coverage and reducing power consumption. To tackle these challenges, using unmanned aerial vehicles (UAVs) would be a good choice. This paper proposes a scheme for uplink non-orthogonal multiple access (NOMA) in UAV communication systems in the presence of granted and grant-free users. At first, the service area users, including granted and grant-free users, are partitioned into some clusters. We propose that the hover location for each cluster is determined considering the weighted mean of users’ locations. We aim to allocate transmission power and form NOMA pairs to maximize the energy efficiency in each cluster subject to the constraints on spectral efficiency and total transmission power. To this end, the transmission powers of each possible pair are obtained, and then Hungarian matching is used to select the best pairs. Finally, finding the flight path of the UAV is modeled by the traveling salesman problem (TSP), and the genetic algorithm method obtains its solution. The results show that the increasing height of the UAV and density of users increases the spectral and energy efficiencies and reduces the outage probability. Also, considering the quality of service (QoS) of granted users for determining the UAV's hover location enhances the transmission's performance.
With the rapid deployment of users and increasing demands for mobile data, communication networks with high capacity are needed more than ever. Furthermore, there are several challenges, such as providing efficient coverage and reducing power consumption. To tackle these challenges, using unmanned aerial vehicles (UAVs) would be a good choice. This paper proposes a scheme for uplink non-orthogonal multiple access (NOMA) in UAV communication systems in the presence of granted and grant-free users. At first, the service area users, including granted and grant-free users, are partitioned into some clusters. We propose that the hover location for each cluster is determined considering the weighted mean of users’ locations. We aim to allocate transmission power and form NOMA pairs to maximize the energy efficiency in each cluster subject to the constraints on spectral efficiency and total transmission power. To this end, the transmission powers of each possible pair are obtained, and then Hungarian matching is used to select the best pairs. Finally, finding the flight path of the UAV is modeled by the traveling salesman problem (TSP), and the genetic algorithm method obtains its solution. The results show that the increasing height of the UAV and density of users increases the spectral and energy efficiencies and reduces the outage probability. Also, considering the quality of service (QoS) of granted users for determining the UAV's hover location enhances the transmission's performance.
[1] R. S. Stansbury, M. A. Vyas, and T. A. Wilson, "A survey of UAS technologies for command, control, and communication (C3)," in Unmanned Aircraft Systems: Springer, 2008, pp. 61-78.
[2] K. P. Valavanis and G. J. Vachtsevanos, Handbook of unmanned aerial vehicles. Springer, 2015.
[3] H. Luo, S.-C. Chu, X. Wu, Z. Wang, and F. Xu, "Traffic collisions early warning aided by small unmanned aerial vehicle companion," Telecommunication systems, vol. 75, pp. 169-180, 2020.
[4] Y. Li et al., "Air-to-ground 3D channel modeling for UAV based on Gauss-Markov mobile model," AEU-International Journal of Electronics and Communications, vol. 114, p. 152995, 2020.
[5] S. Aggarwal and N. Kumar, "Path planning techniques for unmanned aerial vehicles: A review, solutions, and challenges," Computer Communications, vol. 149, pp. 270-299, 2020.
[6] M. Mozaffari, W. Saad, M. Bennis, Y.-H. Nam, and M. Debbah, "A tutorial on UAVs for wireless networks: Applications, challenges, and open problems," IEEE Communications Surveys & Tutorials, vol. 21, no. 3, pp. 2334-2360, 2019.
[7] J. H. Sarker and A. M. Nahhas, "A secure wireless mission critical networking system for unmanned aerial vehicle communications," Telecommunication Systems, vol. 69, no. 2, pp. 237-251, 2018.
[8] S. Sudhakar, V. Vijayakumar, C. S. Kumar, V. Priya, L. Ravi, and V. Subramaniyaswamy, "Unmanned Aerial Vehicle (UAV) based Forest Fire Detection and monitoring for reducing false alarms in forest-fires," Computer Communications, vol. 149, pp. 1-16, 2020.
[9] Q. Liu et al., "Joint power and time allocation in energy harvesting of UAV operating system," Computer Communications, vol. 150, pp. 811-817, 2020.
[10] Y. Zeng, J. Xu, and R. Zhang, "Energy minimization for wireless communication with rotary-wing UAV," IEEE Transactions on Wireless Communications, vol. 18, no. 4, pp. 2329-2345, 2019.
[11] Y. Zeng and R. Zhang, "Energy-efficient UAV communication with trajectory optimization," IEEE Transactions on Wireless Communications, vol. 16, no. 6, pp. 3747-3760, 2017.
[12] M. Hua, Y. Wang, C. Li, Y. Huang, and L. Yang, "Energy-efficient optimization for UAV-aided cellular offloading," IEEE Wireless Communications Letters, vol. 8, no. 3, pp. 769-772, 2019.
[13] J. Yu, R. Zhang, Y. Gao, and L.-L. Yang, "Modularity-based dynamic clustering for energy efficient UAVs-aided communications," IEEE Wireless Communications Letters, vol. 7, no. 5, pp. 728-731, 2018.
[14] Y. Cai, Z. Wei, R. Li, D. W. K. Ng, and J. Yuan, "Energy-efficient resource allocation for secure UAV communication systems," in 2019 IEEE Wireless Communications and Networking Conference (WCNC), 2019: IEEE, pp. 1-8.
[15] M.-N. Nguyen, L. D. Nguyen, T. Q. Duong, and H. D. Tuan, "Real-time optimal resource allocation for embedded UAV communication systems," IEEE Wireless Communications Letters, vol. 8, no. 1, pp. 225-228, 2018.
[16] Y. Chen, W. Feng, and G. Zheng, "Optimum placement of UAV as relays," IEEE Communications Letters, vol. 22, no. 2, pp. 248-251, 2017.
[17] M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, "Wireless communication using unmanned aerial vehicles (UAVs): Optimal transport theory for hover time optimization," IEEE Transactions on Wireless Communications, vol. 16, no. 12, pp. 8052-8066, 2017.
[18] M. Alzenad, A. El-Keyi, F. Lagum, and H. Yanikomeroglu, "3-D placement of an unmanned aerial vehicle base station (UAV-BS) for energy-efficient maximal coverage," IEEE Wireless Communications Letters, vol. 6, no. 4, pp. 434-437, 2017.
[19] R. Amorim et al., "Measured uplink interference caused by aerial vehicles in LTE cellular networks," IEEE Wireless Communications Letters, vol. 7, no. 6, pp. 958-961, 2018.
[20] J. Gong, T.-H. Chang, C. Shen, and X. Chen, "Flight time minimization of UAV for data collection over wireless sensor networks," IEEE Journal on Selected Areas in Communications, vol. 36, no. 9, pp. 1942-1954, 2018.
[21] D. Yang, Q. Wu, Y. Zeng, and R. Zhang, "Energy tradeoff in ground-to-UAV communication via trajectory design," IEEE Transactions on Vehicular Technology, vol. 67, no. 7, pp. 6721-6726, 2018. [22] X. Liu, M. Chen, and C. Yin, "Optimized trajectory design in UAV based cellular networks for 3D users: A double Q-learning approach," 2019. [23] D. Zhai, R. Zhang, L. Cai, B. Li, and Y. Jiang, "Energy-efficient user scheduling and power allocation for NOMA-based wireless networks with massive IoT devices," IEEE Internet of Things Journal, vol. 5, no. 3, pp. 1857-1868, 2018.
[24] M. S. Ali, H. Tabassum, and E. Hossain, "Dynamic user clustering and power allocation for uplink and downlink non-orthogonal multiple access (NOMA) systems," IEEE access, vol. 4, pp. 6325-6343, 2016.
[25] J. Cui, Z. Ding, P. Fan, and N. Al-Dhahir, "Unsupervised machine learning-based user clustering in millimeter-wave-NOMA systems," IEEE Transactions on Wireless Communications, vol. 17, no. 11, pp. 7425-7440, 2018.
[26] M. Zeng, A. Yadav, O. A. Dobre, and H. V. Poor, "Energy-efficient joint user-RB association and power allocation for uplink hybrid NOMA-OMA," IEEE Internet of Things Journal, vol. 6, no. 3, pp. 5119-5131, 2019.
[27] S. Dhakal, P. A. Martin, and P. J. Smith, "NOMA with guaranteed weak user QoS: design and analysis," IEEE Access, vol. 7, pp. 32884-32896, 2019.
[28] X. Mu, Y. Liu, L. Guo, and J. Lin, "Uplink Non-Orthogonal Multiple Access for UAV Communications," CoRR, 2019.
[29] W. Mei and R. Zhang, "Uplink cooperative NOMA for cellular-connected UAV," IEEE Journal of Selected Topics in Signal Processing, vol. 13, no. 3, pp. 644-656, 2019.
[30] R. Duan, J. Wang, C. Jiang, H. Yao, Y. Ren, and Y. Qian, "Resource allocation for multi-UAV aided IoT NOMA uplink transmission systems," IEEE Internet of Things Journal, vol. 6, no. 4, pp. 7025-7037, 2019.
[31] Y. Liu, Z. Qin, Y. Cai, Y. Gao, G. Y. Li, and A. Nallanathan, "UAV communications based on non-orthogonal multiple access," IEEE Wireless Communications, vol. 26, no. 1, pp. 52-57, 2019.
[32] M. Yang, B. Li, Z. Bai, and Z. Yan, "SGMA: Semi-granted multiple access for non-orthogonal multiple access (NOMA) in 5G networking," Journal of Network and Computer Applications, vol. 112, pp. 115-125, 2018.
[33] Z. Ding, R. Schober, P. Fan, and H. V. Poor, "Simple semi-grant-free transmission strategies assisted by non-orthogonal multiple access," IEEE Transactions on Communications, vol. 67, no. 6, pp. 4464-4478, 2019.
[34] Q. Zhang, M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, "Machine learning for predictive on-demand deployment of UAVs for wireless communications," in 2018 IEEE Global Communications Conference (GLOBECOM), 2018: IEEE, pp. 1-6.
[35] M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, "Optimal transport theory for power-efficient deployment of unmanned aerial vehicles," in 2016 IEEE international conference on communications (ICC), 2016: IEEE, pp. 1-6.
[36] H. Tabassum, M. S. Ali, E. Hossain, M. J. Hossain, and D. I. Kim, "Uplink vs. downlink NOMA in cellular networks: Challenges and research directions," in 2017 IEEE 85th vehicular technology conference (VTC Spring), 2017: IEEE, pp. 1-7.
[37] D.-T. Do and M.-S. Van Nguyen, "Outage probability and ergodic capacity analysis of uplink NOMA cellular network with and without interference from D2D pair," Physical Communication, vol. 37, p. 100898, 2019.
[38] A. Hussain, Y. S. Muhammad, M. Nauman Sajid, I. Hussain, A. Mohamd Shoukry, and S. Gani, "Genetic algorithm for traveling salesman problem with modified cycle crossover operator," Computational intelligence and neuroscience, vol. 2017, 2017.
http://jist.acecr.org ISSN 2322-1437 / EISSN:2345-2773 |
Journal of Information Systems and Telecommunication
|
Energy-Efficient User Pairing and Power Allocation for Granted Uplink-NOMA in UAV Communication Systems |
Seyyed Hadi Mostafavi-Amjad1, Vahid Solouk2, Hashem Kalbkhani1*
|
1. Faculty of Electrical Engineering, Urmia University of Technology, Urmia, Iran 2. Department of IT and Computer Engineering, Urmia University of Technology, Urmia, Iran |
Received: 19 Sep 2021/ Revised: 04 Nov 2021/ Accepted: 07 Dec 2021 |
|
Abstract
With the rapid deployment of users and increasing demands for mobile data, communication networks with high capacity are needed more than ever. Furthermore, there are several challenges, such as providing efficient coverage and reducing power consumption. To tackle these challenges, using unmanned aerial vehicles (UAVs) would be a good choice. This paper proposes a scheme for uplink non-orthogonal multiple access (NOMA) in UAV communication systems in the presence of granted and grant-free users. At first, the service area users, including granted and grant-free users, are partitioned into some clusters. We propose that the hover location for each cluster is determined considering the weighted mean of users’ locations. We aim to allocate transmission power and form NOMA pairs to maximize the energy efficiency in each cluster subject to the constraints on spectral efficiency and total transmission power. To this end, the transmission powers of each possible pair are obtained, and then Hungarian matching is used to select the best pairs. Finally, finding the flight path of the UAV is modeled by the traveling salesman problem (TSP), and the genetic algorithm method obtains its solution. The results show that the increasing height of the UAV and density of users increases the spectral and energy efficiencies and reduces the outage probability. Also, considering the quality of service (QoS) of granted users for determining the UAV's hover location enhances the transmission's performance.
Keywords: Energy Efficiency; NOMA; Power Allocation, Unmanned Aerial Vehicle (UAV), Uplink, Users Pairing.
1- Introduction
1-1- Motivation
Nowadays, power consumption is an important challenge in wireless networks, and saving users' energy to achieve high performance is essential. Since users' quality of service (QoS) should be satisfied, energy-saving is more challenging in wireless networks. In future wireless communication networks, particularly in the 5th generation (5G) of wireless communications and beyond, the application of unmanned aerial vehicles (UAVs) is proposed for operating as moving aerial relay nodes or moving aerial base stations (ABSs). UAVs, known as drones in a common tongue, has been the subject of a bunch of research over the past few years [1-5]. If they are fine established and well-operated, UAVs can provide reliable and cost-effective wireless communication solutions for many kinds of real-world scenarios [6]. UAVs can operate better and flexibly than the traditional relay nodes in dense areas. Considering them as ABS has several challenges: power consumption, handover management, channel modeling, low-latency control, 3D localization, and interference management [6-9]. Batteries provide UAVs' power in most cases; therefore, power saving is a severe problem in UAV-assisted communications.
1-2- Contributions
This paper investigates the uplink NOMA communication between the ground users, including granted and grant-free and ABS. Our goal is to partition the users into clusters, form NOMA pairs in each cluster, and allocate power to maximize energy efficiency. Each NOMA pair includes one granted and one grant-free user. To this end, at first, UAV flies right into the communication area and simultaneously starts to send a signal periodically to the ground users to get information about their communication conditions that it needs in the next step to clustering the users. After that, power allocation for each possible pair in each cluster is performed to maximize energy efficiency, while the minimum QoS requirements of users should be satisfied. Finally, efficient pairs are the selection by the Hungarian algorithm. The hover locations for clusters are obtained considering the weighted mean of locations of granted users. Also, the flight path of UAV among different clusters is considered a traversal salesman problem (TSP), and its solution is obtained considering the genetic algorithm. In summary, the contributions of this work are as follows:
1) Introducing new ground-ABS uplink transmission scheme over the granted/grant-free ground users and UAV
2) Each NOMA group consists of one granted and one grant-free user to ensure the QoS of granted users
3) Problem formulation to maximize the energy-efficiency of the proposed transmission subject to the limitation on total transmit power and ensuring the QoS of the users in each NOMA group
4) Proposing weighted-mean based on the QoS of the users to determine the hover location of ABS
5) Proposing joint user groping and power allocation that first obtains the transmit power for each possible NOMA pair and then finds the optimal pairs by Hungarian matching algorithm
6) Considering the flight path of UAV as TSP and utilizing the genetic algorithm to solve it.
The rest of this paper is organized as follows. The previous works are reviewed in Section 2. The system model is described in Section 3. The proposed user clustering, NOMA pair forming, and power allocation is presented in Section 4. Simulation results are given in Section 5. Finally, conclusions remarks are provided in Section 6.
2- Previous Works
Here, we review the related researches which consider UAV-assisted wireless communication topics. In [10], a rotary-wing UAV performs the send and collect data task to/from multiple ground users. This research aimed to optimize the total UAV power consumption by minimizing propulsion and data transmission powers while satisfying each ground node's minimum quality-of-service (QoS) requirement in the uplink direction. Energy-efficient UAV communication with a ground terminal in the downlink direction via optimizing the UAV’s trajectory was studied in [11]. The authors aimed to design a new specimen that considers both the throughput and energy consumption of UAV together. Serving cell edge users by UAV and offloading the data from the base station in downlink direction by circle path for UAV was studied in [12]. The goal was to optimize UAVs' resource allocation, user partitioning, and trajectory by maximizing energy efficiency. A new modularity-based dynamic clustering relying on UAVs' modified Louvain method was studied in [13]. The authors aimed to save the transmitted power of mobile devices in the uplink direction by locating the UAVs on the user clusters' centroids. Resource allocation and trajectory design in downlink direction were formulated as an energy-efficient problem in [14], which jointly optimizes the transmit power, user scheduling, and trajectory and velocity of UAV. A real-time resource allocation algorithm for maximizing the energy efficiency in downlink direction by jointly optimizing the energy-harvesting time and power control for the considered device-to-device (D2D) communication embedded with UAV was proposed in [15].
The optimum establishing of UAV as a relay for maximizing the reliability was studied in [16]. The total power loss, outage, and error rate were considered as the reliability parameters, and optimum height was investigated for static and mobile UAVs. However, it would be better when they consider more than one user on the cell edge. In [17], the effective use of flight-time constrained UAVs as aerial ABSs was investigated to provide coverage for ground users. Notably, a novel framework was proposed for optimizing the average number of bits transmitted to users and UAVs’ hover duration. The authors in [18] proposed an optimum placement algorithm for UAVs that maximizes the number of covered users with minimum transmit power. They have detached the UAV located in the vertical dimension from the horizontal dimension, which simplifies the placement problem. The authors [19] characterized UAV-based communication's latency, reliability, and network availability of ultra-reliable and low-latency communications (URLLC). The height of UAVs and the bandwidth allocation were optimized to minimize the required total bandwidth of URLLC for a given density of UAVs. It was shown that the probability of the line-of-sight (LoS) path and the network availability is strictly concave for the distance between the ground user and UAV. The impact of the height of UAVs connected to the cellular network in uplink was studied in [19]. In [20, 21], a UAV was considered to collect data from a set of sensors with a fixed location. The goal was to minimize the UAV’s total flight time while each sensor could successfully upload its data using a given amount of energy. The problem of trajectory design for UAVs to maximize satisfied users was studied in [22].
Using non-orthogonal multiple access (NOMA) has its benefits and challenges in comparison with other multiple access techniques such as orthogonal-frequency-division-multiple-access (OFDMA) or orthogonal multiple access (OMA). Saving bandwidth by pairing strong and weak users in the same time slots is the most beneficial of NOMA. However, this pairing causes intra-time slot and intra-cell interference challenges. To tackle these challenges, the following works try to solve the problems. To facilitate the serving ground users in a cell, user clustering is a crucial element. Hence, dynamic user scheduling and power allocation problem was proposed in [23] to coordinate the intra-cell interference by minimizing the total power consumption. In [24], sum-rate maximization for uplink and downlink NOMA under the constraints of transmission power limitation, minimum rate requirements of users, and operation constraints were formulated. Machine learning-based user clustering and power allocation algorithms for mmWave-NOMA transmission were considered in [25]. Energy-efficient resource allocation for the uplink of hybrid NOMA and OMA transmission was considered in [26], obtained by jointly optimizing the user clustering, channel assignment, and power allocation. High-rate NOMA, where multiple users share a single zero-forcing beamforming vector, was proposed in [27]. The QoS of all clustered users was satisfied to maintain fairness among the users. In [28], ground-aerial uplink-NOMA of cellular networks was investigated, where ground base stations serve a UAV user and multiple ground users. They aimed to minimize the UAV mission completion time by jointly optimizing the UAV trajectory and association order while considering the UAV’s interference to non-associated ground base stations. In [29], applying of NOMA technique to UAV to cellular BSs uplink communication, under the spectrum sharing with the existing ground users was investigated, and a new cooperative NOMA scheme was proposed to reduce the intense uplink interference due to the UAV’s LoS channels with ground BSs in cellular-connected UAV communication. A combination of multi-UAV communication and NOMA was proposed in [30] to construct the high capacity uplink for the internet of things (IoT) which was achieved by jointly optimizing the sub-channel assignment, transmit power, and flying heights of UAVs. A novel framework for UAV networks with massive access capability supported by NOMA was proposed in [31].
User grouping into two sets to achieve low-latency access and reduce signaling overhead was investigated in [32], where the scheduled-access and random-access users are considered granted and grant-free users, respectively. In [33], NOMA-assisted semi-grant-free transmission was studied, which is investigated compromise between grant-free and grant-based users.
3- System Model
3-1- User Distribution and Transmission Model
As shown in Fig.1, it is assumed that 
granted and 
grant-free terrestrial users are represented by 
and 
respectively. According to the spatial Poisson point process (SPPP) distribution with density 
, these users are distributed in the service area. The locations of the granted user 
and grant-free user 
in two-dimensional space are respectively denoted by 
and 
. These users are partitioned into 
clusters 
. The sets of granted and grant-free users belonging to the cluster 
are given by 
and 
respectively. The number of granted and grant-free users in the cluster is given by 
and 
respectively, therefore we have 
and 
Suppose that 
is the binary matrices with the size of 
If the granted user belongs to the cluster , we have 
otherwise 
On the other side, 
denotes that grant-free user is considered in the cluster , else 
.
Considering Fig. 1, granted user and grant-free user belonging to cluster form the two-user NOMA group 
to transmit their data in uplink direction to ABS. The matrix 
, which presents the pairing of users in the cluster , has the size of 
. If the granted user and grant-free user form the two-user NOMA group in the cluster , then 
, otherwise 
Fig. 1. The system and transmission model that considered in this paper.
The UAV has a hover for a specific time to collect data of the users of each cluster. Since the granted users have priority over the grant-free users, we propose that the hover location is determined based on the weighted mean of granted users in the cluster. Suppose that the hover location of UAV in three-dimensional space for the cluster is denoted by 
, 
, where 
is the height of UAV, then we have
where 
denotes the weight of the granted user for the cluster 
, which is obtained based on the normalized QoS of the user. Since each user only belongs to one cluster, 
only for one cluster, and it is zero for other clusters. The minimum rate requirement of the users is considered as QoS in this paper. Let 
and 
respectively denote the QoS of granted and grant-free users in terms of spectral efficiency in bit/sec/Hz. Then, the weight is computed as:
3-1- Channel Model
Given the granted user located at and the ABS located at , the path loss between the ABS and user will be [34]:
where 
is the carrier frequency in Hz and 
is the speed of light in m/s, and also 
is the Euclidean distance between UAV and user in meter calculated as:
(5)
Also, 
is the normally distributed additional loss depending on environment conditions and distributed as 
and 
for the LoS and NLoS links, respectively. The probability of having an LoS link between the user and ABS at hover location 
is obtained as:
(6)
where 
and 
are constants, and they change depending on the environment and 
is the elevation angle of ABS for the desired user . Then, the probability of having an NLoS link is 
[35]. Eventually, the total path loss from granted user ABS to 
, is obtained as
A similar procedure presented in equations (4)-(7) can be used to obtain the path loss of grant-free user , i.e., 
.
3-2- Spectral and Energy Efficiency
In two-user NOMA transmission, the user with the higher channel gain is called the strong user, and another one is the weak user. The transmitted signal of these users experiences distinct channel gains. In uplink two-user NOMA, the received signal at location k of ABS due to the transmission of pair , i.e., 
, can be obtained as [36]:
(8)
where 
and 
denote the transmitted signal from and , respectively. 
and 
respectively signify the transmit power of granted user and grant-free user . 
and 
represent the channel gain from and to ABS located at kth location, and 
is the white noise with power spectral density 
. The results reported in [36] show that if there is enough separation between weak and strong users, the spectral efficiency of the strong user outperforms that of the weak user. Since satisfying the QoS of granted users has priority over grant-free users, granted users are considered strong users, and grant-free users transmit their data as weak users in each NOMA group.
Transmission of grant-free users interferes with the transmission of granted users and reduces the SNR of the granted user. On the other side, the transmission of the grant-free user receives zero interference from the transmission of the granted user during successive interference cancellation (SIC). Hence, the SINR of the transmission of pair in ABS located in kth location is obtained as follows [36]:
(9)
(10)
Consequently, the SE of NOMA transmission of pair at the kth location of ABS is obtained as:
bits/sec/Hz (11)
With given the spectral efficiency of the pair , the energy efficiency is calculated as follows:
bits/sec/Hz/J (12)
where 
is the total spectral efficiency, 
is the total power consumption, which is the sum of transmit power 
, circuit powers of terrestrial users (
), and power consumed by UAV 
denotes the outage probability. Eventually, described as follow:
(13)
The circuit power of ABS consists of two parts, including the circuit power of ABS in hovering 
and flying times 
The torque coefficient of UAV, 
, is given as [10]:
(14)
Therefore, the corresponding power consumption in hovering time of UAV can be described as [10]:
(15)
and by substitution of in 
, we have [10]:
(16)
The required power for the flight time of rotary-wing UAVs is more intricate than the fixed-wing peer. However, by some mild assumption, the pull coefficient of the blade area is constant, so the torque coefficient for the UAV in flight time with zero climbing angle and speed 
is given as:
(17)
By substituting 
and 
, can be written as a function of forwarding speed and rotor thrust 
as follow:
(18)
Eventually, by definition of torque coefficient, the required power for flight time can be written as follow:
(19)
3-3- Outage Probability
Outage probability is defined as the probability that the SNR or spectral efficiency at the receiver becomes lower than the predefined (or threshold) value. Considering the SIC process, for the strong (or granted) user , the outage probability is obtained as [37]:
(20)
where 
is the probability that the transmit message of the strong user is correctly detected at the receiver, which is calculated as:
(21)
Similarly, for weak (or grant-free) user , we have:
(22)
where the probability of correct detection of the message of the grant-free user is calculated as [37]:
(23)
4- Proposed User Clustering, Power Allocation, and NOMA pair Forming
4-1- Problem Formulation
Over the past decades, energy efficiency has been studied from the information theory perspective. Due to the power limitation of ABS, it is worth considering the energy-efficient transmission scheme and maximizing energy efficiency. Therefore, the proposed scheme for users clustering and joint user pairing and power allocation can be formulated as follows:
subject to:
(S1) |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(S2) |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(S3) |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(S4) |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(S5) |
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(S6) |
|
Algorithm 1. The proposed solution for user clustering, power allocation, and NOMA pair forming |
\\ Cluster forming and obtaining flight path 1. UAV flights over the service area to partition the users to some clusters 2. Compute the hover location for each cluster considering equations (1)-(3) 3. Obtain the flight path by solving TSP via the genetic algorithm \\ Power allocation and NOMA pair forming 4. for each cluster, do 5. Perform power allocation for all possible pairs of granted and grant-free users using equations (32)-(33) 6. Compute energy efficiency of each pair using equation (34) 7. Select the best pairs using the Hungarian algorithm (Algorithm 2) 8. end for |
4-2-1- User Clustering
In order to cluster the terrestrial users, the UAV starts to fly over the service area from a random location. UAV broadcasts the initialization signal and waits to receive the first acknowledgment signal from terrestrial users. This acknowledgment signal contains users' position, desired QoS, and type (granted or grant-free). When the UAV receives the first acknowledgment signal, it starts to create the first cluster. This process continues until the acknowledgment of the first signal is not received in the UAV. At this time, the first cluster is created, and all users whose acknowledgment signal was received are considered in the first cluster. Then, UAV continues to fly over the area to create the second cluster. This process will repeat until all users to be placed in one cluster.
After finishing the clustering process, the UAV computes the hover locations considering equations (1)-(2). In order to minimize the power consumption during fly and hover, the flight path should be minimized. We use TSP integer linear programming with the Dantzig-Fulkerson-Johnson formulation (DFJ) algorithm to obtain the shortest flight path. Suppose that the set 
contains the coordinates of hover locations. TSP solves the following problem:
(26)
subject to:
| (27) | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
(S1) |
| (30) |
(S2) |
| |
(S3) |
| |
(S4) |
| |
(S5) |
|
|
|
|
It is essential to say that the Hungarian method is proper when the matrix is square. Therefore, if the assignment matrix is not square, we must turn it into square form by adding dummy rows or columns. The dummy arrays can be in two forms; they can be equal to the maximum matrix array or be a line with zero numbers; however, zero numbers are recommended. The solution for the Hungarian method is shown in Algorithm 2. The solution of the defined example with the Hungarian matrix method is shown step by step as follow:
(Step 1) Subtract the smallest value in each row from the other values in the row
(Step 2) Each column has zero, so no need to subtract the minimum value from each column.
(Step 1) |
| (Step 2) |
|
|
|
(Step 3) Draw lines through the row and columns that have the 0 entries such that the fewest possible lines are drawn. There are four lines drawn, which is equal to the matrix dimension, so there is the optimal number of zeroes.
Algorithm. 2. The Hungarian algorithm using an adjacency matrix |
1. Convert all the arrays into the reciprocal form 2. if the number of rows and columns are not equal, then 3. Add dummy rows or columns to square the matrix 4. Subtract the smallest entry in each row from all the other entries in the row 5. if there is any column without zero, then 6. Subtract the smallest entry in each column from all the other entries in the column 7. Cover the rows and columns that have the 0 entries with the fewest lines possible are drawn 8. if there the number of lines drawn is equal to the number of rows, then 9. An optimal assignment of zeros is possible, and the algorithm is finished. 10. else if the number of lines is less than number of rows, then 11. The optimal number of zeroes is not yet reached. 12. Go to the next steps. 13. Find the smallest entry not covered by any line. 14. Subtract this entry from each row that is not crossed out, 15. Then add it to each column that is crossed out. 16. end |
(Step 4) Highlight the selected zeros
(Step 3) |
| (Step 4) |
|
|
|
(Step 5) Replace the original values. (Step 6) Replace the primary values of energy efficiency to get which users can be optimally pairs. As shown, there are four lines drawn, and it is equal to the dimension of the matrix, so the algorithm is finished optimally. However, if there are drawn lines less than the matrix dimension, it should follow the algorithm's rules. The Hungarian method, which is shown in Algorithm 2, forms pairs of users optimally. To the extent, one granted user should pair with a grant-free user, which are given as strong and weak users.
(Step 5) |
| (Step 6) |
|
|
|
5- Simulation Results
5-1- Simulation Setup
This section provides numerical results to evaluate the performance of the proposed user clustering, power allocation, and NOMA pair forming to maximize energy efficiency. The UAV acts as a flying ABS in the simulation area and serves the users randomly distributed according to SPPP with density . The parameters used in simulations are given in Table 1.
Table 1. Parameters used in simulations
Parameter | Value |
Simulation area | 5×5 km |
Density of users | (1~3)×10-4 |
Height of ABS | 100 ~ 800 m |
Maximum total transmission power | 23 dBm |
The minimum acceptable received power | -90 dBm |
Noise power | -130 dBm |
Carrier frequency | 1.2 GHz |
Minimum acceptable SNR of granted users | [2 8] dB |
Minimum acceptable SNR of grant-free users | [1 3] dB |
Results in terms of spectral efficiency, energy efficiency, and outage probability are obtained for each pair of UAV height and density of users. For each pair, we run Monte Carlo simulations for 105 trials, and in each trial, the users' locations are generated using SPPP with specific density. Finally, results were averaged.
5-2- Spectral Efficiency
The spectral efficiency for different heights of UAV and density of users is given in Fig. 2 for the total transmission power of 23 dBm. It is observed that increasing the height of UAV and density of users increases the spectral efficiency. In traditional 2D wireless networks such as cellular networks, path loss increases as the distance of users from the base station increases. However, in UAV networks, increasing distance will not necessarily increase the path loss because the probability of the LoS link increases. Increasing the probability of LoS link reduces 
in path loss; hence, overall path loss reduces, SNR increases, and spectral efficiency increases. On the other side, increasing the density of users increases the number of users; therefore, there are more candidate users to form NOMA pairs with better channel conditions. As each pair transmits its data in a specific time slot, increasing the number of users does not increase the interference, and spectral efficiency increases. In summary, for the constant density of users, increasing the height of ABS enhances the spectral efficiency by reducing the path, resulting in higher SNR. For the constant height of ABS, increasing the density of users enhances the spectral efficiency by constructing the pairs with higher SNRs.
Fig. 2. Spectral efficiency for different densities of users and height of UAV.
Fig. 3 shows the impact of hover location on spectral efficiency. Three schemes are utilized to determine the hover location; 1) proposed weighted-mean (WM) of granted users, 2) equal-weight mean (EWM), where we simply consider the mean of the location of the granted user, and 3) random in which UAV randomly hover in the area of the cluster. It is observed that the proposed WM method achieves higher spectral efficiency than the other schemes. Proposed WM determines the hover location near granted users with higher QoS requirements, achieving higher spectral efficiency.
Fig. 3. The effect of hover location on the spectral efficiency
5-3- Energy Efficiency
In Fig. 4, the energy efficiency is given for several heights of UAV and the density of users. It is observed that similar to spectral efficiency, increasing the height of UAV and density of users increases the energy efficiency. Reducing path loss by increasing height reduces the transmit power to satisfy the spectral efficiency; hence energy efficiency increases. On the other side, increasing the density of users provides more candidates for NOMA pairs, which reduces the transmission power and increases energy efficiency. Also, Fig. 5 compares the energy efficiencies obtained for different hover locations. It is observed that the proposed WM scheme outperforms the other schemes considerably.
Fig. 4. Energy efficiency for different densities of users and heights of UAV.
5-4- Outage Probability
It was mentioned that increasing the height of the UAV enhances the uplink transmission by reducing the path loss; therefore, it is expected that outage probability reduces by increasing the height of the UAV, which is depicted in Fig. 6. As spectral and energy efficiencies, outage probability enhances by increasing the height of ABS location. Also, increasing the density of users reduces the outage probability. Fig. 7 compares the outage probability for different hover locations. As expected, the proposed WM scheme has the lowest outage probability since it determines the hover location of the UAV, considering the weighted mean of users’ locations based on their QoS. This approach reduces UAV distance from the users with high QoS requirement and increases their SNR, which reduces the outage probability. Also, increasing the height of the UAV reduces the path loss by increasing the probability of LoS link resulting in lower outage probability.
Fig. 5. The effect of hover location on the energy efficiency
Fig. 6. Outage probability of network for UAV's height.
Fig. 7. The effect of hover location on the outage probability
5-5- Effect of Total Transmission Power
5-6- Comparing Genetic Algorithm with PSO
|
Fig. 8. Effect of total transmission power on energy and spectral efficiencies. |
|
Fig. 9. The effect of the flight path on the energy efficiency |
6- Conclusion
In this paper, the challenges of power allocation and NOMA form pairing in the uplink direction of UAV communication systems were investigated. Users in the UAV coverage area were divided into two classes: granted and grant-free, based on prioritizing the type of demands. Granted and grant-free users are respectively considered as strong and weak users in the NOMA pair. The main criterion for the stated challenges has been to maximize energy efficiency. The optimization problem was formulated to maximize the energy efficiency of transmission subject to the constraints on the minimum acceptable spectral efficiency and total transmission power. In order to solve the problem, at first, transmission powers were computed for each possible NOMA pair, and then, the Hungarian algorithm was employed to select the optimum pairs. The flight path of the UAV was modeled as TSP. The results demonstrated that increasing the height of ABS enhances spectral efficiency, energy efficiency, and outage probability by reducing path loss. Also, increasing the density of users enhances the performance metrics. We also demonstrate that the hover location greatly impacts the performance metrics, and the proposed weighted-mean location outperforms the random and equal-weight men locations.
As future work, we can consider the methods based on machine learning, such as deep belief networks (DBN) for power allocation and pair forming. We can also consider the NOMA clusters with more than two users to support grant-free users in each time slot. Considering different heights for each cluster can be considered as another future work.
Conflicts of Interest
All authors certify that they have NO affiliations with or involvement in any organization or entity with any financial interest (such as honoraria; educational grants; participation in speakers’ bureaus; membership, employment, consultancies, stock ownership, or other equity interest; and expert testimony or patent-licensing arrangements), or non-financial interest (such as personal or professional relationships, affiliations, knowledge or beliefs) in the subject matter or materials discussed in this manuscript.
References
[1] R. S. Stansbury, M. A. Vyas, and T. A. Wilson, "A survey of UAS technologies for command, control, and communication (C3)," in Unmanned Aircraft Systems: Springer, 2008, pp. 61-78.
[2] K. P. Valavanis and G. J. Vachtsevanos, Handbook of unmanned aerial vehicles. Springer, 2015.
[3] H. Luo, S.-C. Chu, X. Wu, Z. Wang, and F. Xu, "Traffic collisions early warning aided by small unmanned aerial vehicle companion," Telecommunication systems, vol. 75, pp. 169-180, 2020.
[4] Y. Li et al., "Air-to-ground 3D channel modeling for UAV based on Gauss-Markov mobile model," AEU-International Journal of Electronics and Communications, vol. 114, p. 152995, 2020.
[5] S. Aggarwal and N. Kumar, "Path planning techniques for unmanned aerial vehicles: A review, solutions, and challenges," Computer Communications, vol. 149, pp. 270-299, 2020.
[6] M. Mozaffari, W. Saad, M. Bennis, Y.-H. Nam, and M. Debbah, "A tutorial on UAVs for wireless networks: Applications, challenges, and open problems," IEEE Communications Surveys & Tutorials, vol. 21, no. 3, pp. 2334-2360, 2019.
[7] J. H. Sarker and A. M. Nahhas, "A secure wireless mission critical networking system for unmanned aerial vehicle communications," Telecommunication Systems, vol. 69, no. 2, pp. 237-251, 2018.
[8] S. Sudhakar, V. Vijayakumar, C. S. Kumar, V. Priya, L. Ravi, and V. Subramaniyaswamy, "Unmanned Aerial Vehicle (UAV) based Forest Fire Detection and monitoring for reducing false alarms in forest-fires," Computer Communications, vol. 149, pp. 1-16, 2020.
[9] Q. Liu et al., "Joint power and time allocation in energy harvesting of UAV operating system," Computer Communications, vol. 150, pp. 811-817, 2020.
[10] Y. Zeng, J. Xu, and R. Zhang, "Energy minimization for wireless communication with rotary-wing UAV," IEEE Transactions on Wireless Communications, vol. 18, no. 4, pp. 2329-2345, 2019.
[11] Y. Zeng and R. Zhang, "Energy-efficient UAV communication with trajectory optimization," IEEE Transactions on Wireless Communications, vol. 16, no. 6, pp. 3747-3760, 2017.
[12] M. Hua, Y. Wang, C. Li, Y. Huang, and L. Yang, "Energy-efficient optimization for UAV-aided cellular offloading," IEEE Wireless Communications Letters, vol. 8, no. 3, pp. 769-772, 2019.
[13] J. Yu, R. Zhang, Y. Gao, and L.-L. Yang, "Modularity-based dynamic clustering for energy efficient UAVs-aided communications," IEEE Wireless Communications Letters, vol. 7, no. 5, pp. 728-731, 2018.
[14] Y. Cai, Z. Wei, R. Li, D. W. K. Ng, and J. Yuan, "Energy-efficient resource allocation for secure UAV communication systems," in 2019 IEEE Wireless Communications and Networking Conference (WCNC), 2019: IEEE, pp. 1-8.
[15] M.-N. Nguyen, L. D. Nguyen, T. Q. Duong, and H. D. Tuan, "Real-time optimal resource allocation for embedded UAV communication systems," IEEE Wireless Communications Letters, vol. 8, no. 1, pp. 225-228, 2018.
[16] Y. Chen, W. Feng, and G. Zheng, "Optimum placement of UAV as relays," IEEE Communications Letters, vol. 22, no. 2, pp. 248-251, 2017.
[17] M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, "Wireless communication using unmanned aerial vehicles (UAVs): Optimal transport theory for hover time optimization," IEEE Transactions on Wireless Communications, vol. 16, no. 12, pp. 8052-8066, 2017.
[18] M. Alzenad, A. El-Keyi, F. Lagum, and H. Yanikomeroglu, "3-D placement of an unmanned aerial vehicle base station (UAV-BS) for energy-efficient maximal coverage," IEEE Wireless Communications Letters, vol. 6, no. 4, pp. 434-437, 2017.
[19] R. Amorim et al., "Measured uplink interference caused by aerial vehicles in LTE cellular networks," IEEE Wireless Communications Letters, vol. 7, no. 6, pp. 958-961, 2018.
[20] J. Gong, T.-H. Chang, C. Shen, and X. Chen, "Flight time minimization of UAV for data collection over wireless sensor networks," IEEE Journal on Selected Areas in Communications, vol. 36, no. 9, pp. 1942-1954, 2018.
[21] D. Yang, Q. Wu, Y. Zeng, and R. Zhang, "Energy tradeoff in ground-to-UAV communication via trajectory design," IEEE Transactions on Vehicular Technology, vol. 67, no. 7, pp. 6721-6726, 2018.
[22] X. Liu, M. Chen, and C. Yin, "Optimized trajectory design in UAV based cellular networks for 3D users: A double Q-learning approach," 2019.
[23] D. Zhai, R. Zhang, L. Cai, B. Li, and Y. Jiang, "Energy-efficient user scheduling and power allocation for NOMA-based wireless networks with massive IoT devices," IEEE Internet of Things Journal, vol. 5, no. 3, pp. 1857-1868, 2018.
[24] M. S. Ali, H. Tabassum, and E. Hossain, "Dynamic user clustering and power allocation for uplink and downlink non-orthogonal multiple access (NOMA) systems," IEEE access, vol. 4, pp. 6325-6343, 2016.
[25] J. Cui, Z. Ding, P. Fan, and N. Al-Dhahir, "Unsupervised machine learning-based user clustering in millimeter-wave-NOMA systems," IEEE Transactions on Wireless Communications, vol. 17, no. 11, pp. 7425-7440, 2018.
[26] M. Zeng, A. Yadav, O. A. Dobre, and H. V. Poor, "Energy-efficient joint user-RB association and power allocation for uplink hybrid NOMA-OMA," IEEE Internet of Things Journal, vol. 6, no. 3, pp. 5119-5131, 2019.
[27] S. Dhakal, P. A. Martin, and P. J. Smith, "NOMA with guaranteed weak user QoS: design and analysis," IEEE Access, vol. 7, pp. 32884-32896, 2019.
[28] X. Mu, Y. Liu, L. Guo, and J. Lin, "Uplink Non-Orthogonal Multiple Access for UAV Communications," CoRR, 2019.
[29] W. Mei and R. Zhang, "Uplink cooperative NOMA for cellular-connected UAV," IEEE Journal of Selected Topics in Signal Processing, vol. 13, no. 3, pp. 644-656, 2019.
[30] R. Duan, J. Wang, C. Jiang, H. Yao, Y. Ren, and Y. Qian, "Resource allocation for multi-UAV aided IoT NOMA uplink transmission systems," IEEE Internet of Things Journal, vol. 6, no. 4, pp. 7025-7037, 2019.
[31] Y. Liu, Z. Qin, Y. Cai, Y. Gao, G. Y. Li, and A. Nallanathan, "UAV communications based on non-orthogonal multiple access," IEEE Wireless Communications, vol. 26, no. 1, pp. 52-57, 2019.
[32] M. Yang, B. Li, Z. Bai, and Z. Yan, "SGMA: Semi-granted multiple access for non-orthogonal multiple access (NOMA) in 5G networking," Journal of Network and Computer Applications, vol. 112, pp. 115-125, 2018.
[33] Z. Ding, R. Schober, P. Fan, and H. V. Poor, "Simple semi-grant-free transmission strategies assisted by non-orthogonal multiple access," IEEE Transactions on Communications, vol. 67, no. 6, pp. 4464-4478, 2019.
[34] Q. Zhang, M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, "Machine learning for predictive on-demand deployment of UAVs for wireless communications," in 2018 IEEE Global Communications Conference (GLOBECOM), 2018: IEEE, pp. 1-6.
[35] M. Mozaffari, W. Saad, M. Bennis, and M. Debbah, "Optimal transport theory for power-efficient deployment of unmanned aerial vehicles," in 2016 IEEE international conference on communications (ICC), 2016: IEEE, pp. 1-6.
[36] H. Tabassum, M. S. Ali, E. Hossain, M. J. Hossain, and D. I. Kim, "Uplink vs. downlink NOMA in cellular networks: Challenges and research directions," in 2017 IEEE 85th vehicular technology conference (VTC Spring), 2017: IEEE, pp. 1-7.
[37] D.-T. Do and M.-S. Van Nguyen, "Outage probability and ergodic capacity analysis of uplink NOMA cellular network with and without interference from D2D pair," Physical Communication, vol. 37, p. 100898, 2019.
[38] A. Hussain, Y. S. Muhammad, M. Nauman Sajid, I. Hussain, A. Mohamd Shoukry, and S. Gani, "Genetic algorithm for traveling salesman problem with modified cycle crossover operator," Computational intelligence and neuroscience, vol. 2017, 2017.
* Hashem Kalbkhani
رایمگ
پیوندهای سایت
مراکز مرتبط
پشتیبانی
صفحات رسمی
حقوق این وبسایت متعلق به سامانه مدیریت نشریات رایمگ است.
حق نشر © 1403-1396