Performance Analysis of a SWIPT Enabled UAV-Assisted Relaying

In the recent era, unmanned aerial vehicle (UAV) plays an important role in numerous application fields related to the wireless communication system. Due to its precise control, efficient deployment, and affordable cost, UAV-assisted communication attracts significant attention to all the sectors including the defense sector, agriculture sector, and security purpose, and so on. Though UAVassisted relaying has enormous advantages but there are potential challenges while UAV deploys as a relay. For example, deploying UAV in the wireless communication field, its battery life is the main concern due to its limited battery size and storage capacity. To get significant benefits from UAV while deployed in the cooperative communication network, the battery status of the UAV is an unavoidable issue. To minimize the aforementioned problem, energy harvesting (EH) techniques can be an efficient solution. The UAV can harvest energy from the transmitted power by the source and with the help of this harvested energy UAV can retransmit the signal to the destination. However, there are several parameters that also significantly influence the UAV-based cooperative system performance such as UAV’s position, time allocation factor and power allocation factor, and UAV’s height. Considering the importance of the aforementioned parameters, in this paper, we have considered simultaneous wireless information and power transfer (SWIPT) enabled UAV-assisted relaying network and evaluate the system outage performance with different parameters aspects. We have provided some insight about the parameters such as the UAV’s position, the power allocation factor and the time allocation factor and the UAV’s height by providing simulation results such as the outage probability versus transmit power in the different urban scenario, the outage probability versus time allocation factor and power allocation factor and the outage probability versus UAV’s height. These simulation results clearly show the significance of the abovementioned parameters in wireless-powered UAVassisted cooperative communication.


Introduction
lenges such as networking architectures, channel characteristics, etc [2]. UAV also served as a flying base station to facilitate wireless coverage in the different geographical areas [3]. UAV can also deploy in 5G technology which could be an important component to provide wireless communication networks with enhanced data rate [4]. However, UAV can also play a potential role in the Internet of Things (IoT) to serve small battery-constraint devices such as sensor networks [5].

Related works
At present, numerous research activities are ongoing on UAV-based wireless communication systems. UAV altitude is an important parameter to extend network coverage. In [6], analysis shows that optimum altitude is the function of the statistical parameters of the different urban environments and maximum allowed pathloss. While UAV served as a mobile base station to ensure the coverage to all the ground terminals by placement optimization techniques presented in [7]. Height-dependent path-loss exponent was introduced in [8], which minimizes the system outage probability of an A2G link. Optimum altitudes for both static and mobile UAV are studied in [9]. However, the utilization of multiple UAV can enhance system performance.
The capacity increases as the number of UAV increases proportionally showed in [10]. Considering multi UAV relaying [11], the dual-hop multilink scenario provides better performance than the multi-hop single link. Nevertheless, the UAV's Limited energy storage capacity is one of the main concerns while deploying UAV in a wireless cooperative network. Meanwhile, energy harvesting (EH) can be an efficient solution for this problem. UAV-enabled relaying with energy harvesting techniques in the different urban scenarios has been studied in [12]. Simultaneous Wireless Information and Power Transfer (SWIPT) is now one of the potential editions in wireless communication systems. To transfer energy and information to incorporate between smart communications systems, SWIPT plays an important role including 5G technology, IoT technology and mobile edge computing, and so on. Different SWIPT architecture has been studied in [13] including recent advances and future directions. UAV assisted communication with mmWave SWIPT techniques to provide secrecy performance evaluation studied in [14]. With time splitting (TS) and power splitting (PS) scheme, UAV-assisted decode and forward network were considered to provide closedform expression of outage and BER with aggregate interference showed in [15]. Considering SWIPT technology in [16], they optimized trajectory profile, PS ratio and power profile to maximize throughput. In our previous work [17], we have considered SWIPT enabled UAV to search optimal UAV position for low altitude base UAV's channel. To find the optimal solution with two steps, first of all we optimized the power splitting factor with given TA then we optimized TA with conditionally optimized power splitting factor expressed in the closed form [18].
Considering the above-mentioned work, we can easily relate that several factors need to be taken into account while utilizes UAV as a relay in a cooperative communication system. Such as UAV's battery status as well as the parameters like the UAV's position, the time allocation factor, the power allocation factor, and the UAV's height which has a significant role in the system performance.
Motivated by the aforementioned performance controlling factors, in this paper, we have designed a UAV-assisted wireless communication system with SWIPT technology. We evaluate the system outage performance with various parameters aspects. We evaluate the system outage performance as a function of UAV's position, power allocation factor and time allocation factor as well as UAV's height which is also a key factor for a reliable UAV-based wireless communication system. Simulation results clearly show that the significance of SWIPT technology in UAV-assisted cooperative networks and as well as the important insight of different performance measuring parameters.

System model
We have considered a system illustrated in Fig.  1 with a single base station (S), a mobile station (D) and a UAV which deploys as decode and forward relay. All the nodes are equipped with single antenna. The UAV having altitude H including the ground distance r 1 and r 2 from S, D respectively. The distance between S-to-UAV and UAV-to-D is denoted by d 1 , d 2 respectively. While θ 1 , θ 2 represent the elevation angle between S to UAV and UAV to D respectively. The relationship between the parameters can be expressed as θ i = tan −1 H ri . UAV-to-D is denoted by 1 , 2 respectively. While 1 , 2 represent the elevation angle between S to UAV and UAV to D respectively. The relationship between the parameters can be expressed as = tan −1 .
Where ̅ is the mean and (. ) is the zero order modified Bessel function of first kind [20]. is the Rician factor which is defined as the ratio of the power in LoS and NLoS power in multi-scatters.   The wireless channel between S-UAV and UAV-D experienced small-scale fading and large-scale path-loss. Since the receiver experienced LoS and multipath scattering, Rician distribution is the suitable choice to model smallscale fading. The channel fading power can be modeled as , which can be expressed as [19] f Xi (x) = (k i + 1) where X i is the mean and I o (.) is the zero order modified Bessel function of first kind [20]; K i is the Rician factor which is defined as the ratio of the power in LoS and NLoS power in multiscatters. Figure 2 represents the system relaying protocols where in phase I, source transmits the signal towards the UAV relay which then harvests energy from the received signal and as well as decodes it at τ T time, where τ ∈ (0, 1). With this harvested energy, the UAV transfers the information to the destination with (1 − τ ) T time in the second phase. The power splitting ratio is (1 − α) : α, where α ∈ (0, 1). Without loss of generality, we assume T = 1 in this system. 2 optimized TA with conditionally optimized power splitting factor expressed in the closed form [18].
Considering the above-mentioned work, we can easily relate that several factors need to be taken into account while utilizes UAV as a relay in a cooperative communication system. Such as UAV's battery status as well as the parameters like the UAV's position, the time allocation factor, the power allocation factor, and the UAV's height which has a significant role in the system performance.
Motivated by the aforementioned performance controlling factors, in this paper, we have designed a UAV-assisted wireless communication system with SWIPT technology. We evaluate the system outage performance with various parameters aspects. We evaluate the system outage performance as a function of UAV's position, power allocation factor and time allocation factor as well as UAV's height which is also a key factor for a reliable UAV-based wireless communication system. Simulation results clearly show that the significance of SWIPT technology in UAV-assisted cooperative networks and as well as the important insight of different performance measuring parameters.

III. SYSTEM MODEL
We have Considered a system illustrated in Figure 1 with a single base station (S), a mobile station (D) and a UAV which deploys as decode and forward relay. All the nodes are equipped with single antenna. The UAV having altitude H including the ground distance 1 and 2 from S, D respectively. The distance between S-to-UAV and experienced small-scale fading and large-scale path-loss. Since the receiver experienced LoS and multipath scattering, Rician distribution is the suitable choice to model small-scale fading. The channel fading power can be modeled as = |ℎ | 2 (1) ~ 2 ( ), which can be expressed as [19] ( ) = Where ̅ is the mean and (. ) is the zero order modified Bessel function of first kind [20]. is the Rician factor which is defined as the ratio of the power in LoS and NLoS power in multi-scatters.   The harvested energy at the UAV device can be expressed as where η is the energy harvesting efficiency and P 1 is the transmit power of S. P L 1 is the pathloss between S to UAV which can be expressed as where η LOS1 , η N LOS1 , a 1 , b 1 are the constant parameters belongs to the propagation environment [21]. c = 3 × 10 8 m/s & f = 2 GHz is the speed of light and frequency respectively.
The deliverable rate from S to UAV is expressed as where σ 1 2 is the noise power at UAV.
The deliverable rate from UAV to D is expressed as where σ 2 2 the noise power at D. P L 2 is the pathloss between UAV to D which can be expressed as where η LOS2 , η N LOS2 , a 2 , b 2 are the constant parameters belonging to the propagation environment.

Outage probability analysis
The outage probability of the system considering information rate R t is given by Equation (8) can be further modified as P Out can be represented as (10) where F X2 c Γ2 u is the CCDF of the random variable X 2 which can be given by is the first order Marcum Qfunction [22]. Now, the pdf of the random variable X 1 is given by Now substituting (11) and (12) in (10), the outage probability is finally expressed as

Simulation results
To  Performance Analysis of a SWIPT Enabled UAV-Assisted Relaying 4 represents the monte carlo simulation results which completely matched with each other. We observe the outage probability performance for urban, suburban, dense urban and highrise environment conditions.

Fig. 3.
Outage probability Vs transmits power 1 for different urban environments.
As expected, the optimal result provided by the suburban scenario due to the better LoS condition from other environment scenarios. Next we show the outage probability versus different Rician factor setup to see the system performance. As we all know, the lower 1 , 2 values indicate that a more severe fading scenario which reflects in figure 4. When 1 = 2 = 1 , the outage performance is worst compared to results with higher 1 , 2 scenarios. We also extend our simulation results to observe the outage performance for different transmission data rates. For this simulation we set up the parameters as 1 = 2 = = 10 , 1 = 2 = 4 for the suburban scenario. As we can see from figure 5, as the data rate increases the system outage performance degrades. This phenomenon happens as the system failed to maintain the requirements for higher data rate transmission. Meanwhile, UAV's position ( , 1 ) is the most important parameter which influences the system's performance mostly. To observe the issue, we also represent the outage probability Vs ground distance 1 and the UAV's height . As we can see from the figure 6 which represent the outage probability varies with different UAV location ( , 1 ). From the figure, when the source and the UAV located far from each other, the outage performance is worst for this scenario. It clearly indicates that there must be an optimal UAV's position ( , 1 ) which can minimize the outage of the network. The marker shows the optimal UAV's location ( , 1 ) in figure 6. As expected, the optimal result provided by the suburban scenario due to the better LoS condition from other environment scenarios. Next we show the outage probability versus different Rician factor setup to see the system performance.
As we all know, the lower K 1 , K 2 values indicate that a more severe fading scenario which reflects in Fig. 4. When K 1 = K 2 = 1, the outage performance is worst compared to results with higher K 1 , K 2 scenarios.
Performance Analysis of a SWIPT Enabled UAV-Assisted Relaying 4 represents the monte carlo simulation results which completely matched with each other. We observe the outage probability performance for urban, suburban, dense urban and highrise environment conditions.

Fig. 3.
Outage probability Vs transmits power 1 for different urban environments.
As expected, the optimal result provided by the suburban scenario due to the better LoS condition from other environment scenarios. Next we show the outage probability versus different Rician factor setup to see the system performance. As we all know, the lower 1 , 2 values indicate that a more severe fading scenario which reflects in figure 4. When 1 = 2 = 1 , the outage performance is worst compared to results with higher 1 , 2 scenarios. We also extend our simulation results to observe the outage performance for different transmission data rates. For this simulation we set up the parameters as 1 = 2 = = 10 , 1 = 2 = 4 for the suburban scenario. As we can see from figure 5, as the data rate increases the system outage performance degrades. This phenomenon happens as the system failed to maintain the requirements for higher data rate transmission. Meanwhile, UAV's position ( , 1 ) is the most important parameter which influences the system's performance mostly. To observe the issue, we also represent the outage probability Vs ground distance 1 and the UAV's height . As we can see from the figure 6 which represent the outage probability varies with different UAV location ( , 1 ). From the figure, when the source and the UAV located far from each other, the outage performance is worst for this scenario. It clearly indicates that there must be an optimal UAV's position ( , 1 ) which can minimize the outage of the network. The marker shows the optimal UAV's location ( , 1 ) in figure 6. We also extend our simulation results to observe the outage performance for different trans-mission data rates. For this simulation we set up the parameters as r 1 = r 2 = H = 10 m, K 1 = K 2 = 4 for the suburban scenario. As we can see from Fig. 5, as the data rate increases the system outage performance degrades. This phenomenon happens as the system failed to maintain the requirements for higher data rate transmission.
Performance Analysis of a SWIPT Enabled UAV-Assisted Relaying 4 represents the monte carlo simulation results which completely matched with each other. We observe the outage probability performance for urban, suburban, dense urban and highrise environment conditions.

Fig. 3.
Outage probability Vs transmits power 1 for different urban environments.
As expected, the optimal result provided by the suburban scenario due to the better LoS condition from other environment scenarios. Next we show the outage probability versus different Rician factor setup to see the system performance. As we all know, the lower 1 , 2 values indicate that a more severe fading scenario which reflects in figure 4. When 1 = 2 = 1 , the outage performance is worst compared to results with higher 1 , 2 scenarios. We also extend our simulation results to observe the outage performance for different transmission data rates. For this simulation we set up the parameters as 1 = 2 = = 10 , 1 = 2 = 4 for the suburban scenario. As we can see from figure 5, as the data rate increases the system outage performance degrades. This phenomenon happens as the system failed to maintain the requirements for higher data rate transmission. Meanwhile, UAV's position ( , 1 ) is the most important parameter which influences the system's performance mostly. To observe the issue, we also represent the outage probability Vs ground distance 1 and the UAV's height . As we can see from the figure 6 which represent the outage probability varies with different UAV location ( , 1 ). From the figure, when the source and the UAV located far from each other, the outage performance is worst for this scenario. It clearly indicates that there must be an optimal UAV's position ( , 1 ) which can minimize the outage of the network. The marker shows the optimal UAV's location ( , 1 ) in figure 6. Meanwhile, UAV's position (H, r 1 ) is the most important parameter which influences the system's performance mostly. To observe the issue, we also represent the outage probability Vs ground distance r 1 and the UAV's height H. As we can see from the Fig. 6 which represents the outage probability varies with different UAV location (H, r 1 ). From the figure, when the source and the UAV located far from each other, the outage performance is the worst for this scenario.
It clearly indicates that there must be an optimal UAV's position (H, r 1 ) which can minimize the outage of the network. The marker shows the optimal UAV's location (H, r 1 ) in Fig. 6.
However, other significant system performance parameters are the time allocation factor and the power allocation factor. To evaluate the influence of those parameters, we provide the outage probability Vs the time allocation and the power allocation factor in Fig. 7 in a suburban environment with transmit power P 1 = 40 dBm and k 1 = k 2 = 4. From Fig. 7, we can see that 4 tage the , 2 ich tage her for the tes.

= As
parameter which influences the system's performance mostly. To observe the issue, we also represent the outage probability Vs ground distance 1 and the UAV's height . As we can see from the figure 6 which represent the outage probability varies with different UAV location ( , 1 ). From the figure, when the source and the UAV located far from each other, the outage performance is worst for this scenario. It clearly indicates that there must be an optimal UAV's position ( , 1 ) which can minimize the outage of the network. The marker shows the optimal UAV's location ( , 1 ) in figure 6. when the τ is increasing which means that the UAV gets more time in the harvesting part so that the system performance improves but after a certain value of τ the performance becomes degrade because the UAV failed to retransmit the signal to the destination due to less time allocated to the second phase of transmission. The marker shows the optimal α, τ values set up for this scenario which minimizes system outage. It is obvious that as UAV harvests more energy the system performance will be better. However, other significant system performance parameters are the time allocation factor and the power allocation factor. To evaluate the influence of those parameters, we provide the outage probability Vs the time allocation and the power allocation factor in figure 7 in a suburban environment with transmit power 1 = 40 dBm amd 1 = 2 = 4. From figure 7, we can see that when the is increasing which means that the UAV gets more time in the harvesting part so that the system performance improves but after a certain value of the performance becomes degrade because the UAV failed to retransmit the signal to the destination due to less time allocated to the second phase of transmission. The marker shows the optimal , values set up for this scenario which minimizes system outage. It is obvious that as UAV harvests more energy the system performance will be better. On the other hand, the UAV's height has also a significant impact on the system performance. Considering this issue, we also evaluate the system outage probability as a function of UAV's height in the different urban scenarios. We can easily observe from figure 8 that the outage probability first decreases as UAV's height increases up-to a certain value then the outage probability again increases as UAV's height increases. When the UAV's height is very low, because of the NLoS issue the system outage probability is high. But as the UAV's height starts increasing the outage probability decreases. Then suddenly after a certain height, the outage is again increasing due to a higher path loss which indicates that there exists an optimal UAV's height which minimizes the system outage probability.  On the other hand, the UAV's height has also a significant impact on the system performance.

UAV-assisted wireless communication recently gains potential attraction in recent period for its vast benefits
Considering this issue, we also evaluate the system outage probability as a function of UAV's height in the different urban scenarios. We can easily observe from Fig. 8 that the outage probability first decreases as UAV's height increases up-to a certain value then the outage probability again increases as UAV's height increases. However, other significant system performance parameters are the time allocation factor and the power allocation factor. To evaluate the influence of those parameters, we provide the outage probability Vs the time allocation and the power allocation factor in figure 7 in a suburban environment with transmit power 1 = 40 dBm amd 1 = 2 = 4. From figure 7, we can see that when the is increasing which means that the UAV gets more time in the harvesting part so that the system performance improves but after a certain value of the performance becomes degrade because the UAV failed to retransmit the signal to the destination due to less time allocated to the second phase of transmission. The marker shows the optimal , values set up for this scenario which minimizes system outage. It is obvious that as UAV harvests more energy the system performance will be better. On the other hand, the UAV's height has also a significant impact on the system performance. Considering this issue, we also evaluate the system outage probability as a function of UAV's height in the different urban scenarios. We can easily observe from figure 8 that the outage probability first decreases as UAV's height increases up-to a certain value then the outage probability again increases as UAV's height increases. When the UAV's height is very low, because of the NLoS issue the system outage probability is high. But as the UAV's height starts increasing the outage probability decreases. Then suddenly after a certain height, the outage is again increasing due to a higher path loss which indicates that there exists an optimal UAV's height which minimizes the system outage probability.

VI. CONCLUSIONS
UAV-assisted wireless communication recently gains potential attraction in recent period for its vast benefits and opportunities. However, it provides an enormous advantage where human access is impossible or dangerous. Though the UAV has significant benefits while deploying as a relay but on the other hand the challenges also take into account such as UAV's battery life, precise design, accuracy, etc. Give priority to battery life issues of UAV, numerous researches are ongoing to provide UAVbased communication safer, faster and reliable. Recently SWIPT technology draws significant attention to the researcher. Utilizing SWIPT technology shows major improvement in UAV-assisted wireless communication. While UAV deploys as a relay, parameters such as UAV position, time and power allocation factors and, UAV's height need to be considered. In this paper, we summarize the importance of those parameters with SWIPT technology with different simulation results. Our results indicate the convenience of SWIPT enabled UAV-assisted cooperative communication. When the UAV's height is very low, because of the NLoS issue the system outage probability is high. But as the UAV's height starts increasing the outage probability decreases. Then suddenly after a certain height, the outage is again increasing due to a higher path loss which indicates that there exists an optimal UAV's height which minimizes the system outage probability.

Conclusions
UAV-assisted wireless communication recently gains potential attraction in recent period for its vast benefits and opportunities. However, it provides an enormous advantage where human access is impossible or dangerous. Though the UAV has significant benefits while deploying as a relay but on the other hand the challenges also take into account such as UAV's battery life, precise design, accuracy, etc. Give priority to battery life issues of UAV, numerous researches are ongoing to provide UAV-based communication safer, faster and reliable. Recently SWIPT technology draws significant attention to the researcher. Utilizing SWIPT technology shows major improvement in UAV-assisted wire-less communication. While UAV deploys as a relay, parameters such as UAV position, time and power allocation factors and, UAV's height need to be considered. In this paper, we summarize the importance of those parameters with SWIPT technology with different simulation results. Our results indicate the convenience of SWIPT enabled UAV-assisted cooperative communication.