Modeling A dominant LoS (Line of Sight) component is preferred for the air to air (A2A) communication channels of UAV communication. A2A propagation channel is useful in multi hop UAV networks for sensing and coordination applications. In case of emergency situations, it can replace the existing communication systems and provide a backhaul wireless connectivity. Large scale fading statistics in A2A propagation channels have been discussed in Refs. [40–45]. But the impact of antenna orientation, gain from UAV-MIMO system and Doppler spectrum of the A2A channel are yet to be worked upon. The major challenge encountered by these channels is the higher Doppler frequency shifts which occur due to higher relative velocity between UAVs. In 5G networks, if millimeter waves are used for backhauling than it would result in wider spectrum allocations, higher data rates, reduced latency in A2A channels. But the drawback is higher Doppler shifts. Hence we need to work out suitable techniques to be used in A2A links.
Various empirical and analytical channel models characterizing A2A and A2G propagation channels have been discussed in Ref. [46]. Multidimensional UAV channel modeling is yet to be explored thoroughly.
2.2.2 UAV-Assisted Cellular Network Planning and Provisioning
Network planning is more challenging in the case of UAV assisted cellular network. Parameters such as mobility, line of sight interference, energy constraints, and wireless backhaul connectivity have to be considered while planning such networks. The network needs to be planned taking into consideration the LoS interference between numbers of UAV-UE in the uplink connections. The BS in UAV assisted networks should be able to offer 3D communication coverage as the UAV-UEs are located at heights greater than the conventional BS antenna heights. The types of antennas at the ground stations involved in the UAV-UE in downlink communication need to be redesigned to provide wider coverage over the sky. The characteristics of the channel between UAV and BS are different from the channels of conventional terrestrial systems, as strong LoS links exist in this scenario. Such strong communications links result in efficient communication between UAV and the associated BS but also poses the threat of inter-cell interference from adjacent but non-associated BSs, in scenarios having both aerial and terrestrial UEs [47]. One also needs to consider the signaling and overhead involved in such networks due to the mobility feature of the UAVs. Studies carried out by researchers and presented in Refs. [18, 20, 48–50] focuses on problems like user association, backhaul connectivity, optimization of the number of UAVs that should be deployed in a network, placement of UAVs etc. Researchers have thrived to prove that optimal planning of UAV assisted cellular networks would require exhaustive work to be carried out to achieve an enhancement in throughput, reduction in delay, reduction of signaling overheads, reduction in interference in case of multiple UAV scenarios, less operational cost, reduced energy consumption, better effective interference mitigation techniques to deal with UAV-BS channel, techniques for supporting asymmetric UAV traffic requirements and so on.
2.2.3 Millimeter Wave Cellular Connected UAVs
The challenges encountered by cellular networks operating at mmWaves (Millimeter Waves) can be enumerated as, high attenuation, reduced transmission range, increased scattering, high penetration losses when encountering objects, frequent signal blockage, etc. These can be overcome to a certain extent by the use of UAV-assisted cellular networks. But such UAV based wireless communication systems operating at mmWaves also face certain issues. The channel characteristics of UAV mmWave communication networks are quite different as compared to those of the traditional UAV communication networks as well as the terrestrial cellular network communication. Channel models incorporating air to air channels, air to ground channel, air to sea channels need to be designed for UAV mmWave communication networks. There is need to carry out both empirical as well as analytical studies for UAV mmWave channels used in dense urban scenarios. Need arises to develop models that would assist in study of effects of weather on performance of UAV assisted mmWave networks, as mmWave propagation and the stability of UAV both get affected by rain and wind. The fast channel estimation methods should be developed to solve the problems due to reduced channel coherence time in high-mobility UAV communication systems. At mmWave frequencies more efficient beam training and tracking methods, which can handle rapid variations of path gain and fast deviations in angle of arrival and departure of beam, are required. Strategies to detect the abrupt changes and designing of smart precoders are required to improve the performance of UAV assisted mmWave communication systems. In UAV to BS mmWave communication, UAV detection and positioning is another critical challenge to be sorted out. In order to solve the interference problems, beam width needs to be designed judiciously. Higher frequencies will give narrow beams but result in heavy training overheads for beam alignment as in UAV-to-BS mmWave communication. On other hand, broader beams will increase interference to other cells. Efficient spectrum sharing schemes for increasing the network throughput and spectral efficiency needs to be designed for UAV mmWave communication.
2.2.4 Deployment of UAV
Another challenging issue in UAV based communication networks is the deployment of the UAVs. Parameters like the geographical area, location of ground users, altitude of the UAV, etc. play a critical role in determining the performance of UAV-based communications Simultaneous deployment of UAVs causes inter system interference. Both distance and Line of sight probability are to be considered while deciding the optimal altitude. Deployment at lower altitudes poses problems of lower coverage and less probabilities of LoS links due to the shadowing effect whereas UAVs at higher altitudes tend to exhibit poor coverage performances on account of higher path losses because of the large distance between transmitter and receiver [51]. Research work of Refs. [15, 16, 20, 52] discusses the algorithms developed to find the optimal placement of LAP’s, the maximum number of UAVs required to serve all the users on the ground, the impact of UAV’s altitude on the performance of networks, placement of UAV’s for maximizing the coverage etc. Optimal deployment of UAVs reduces the average transmit power of the devices in an UAV-IoT communication network [53]. Also, the number of UAV’s required in such networks is also determined by the altitude of the UAVs [33]. UAVs with directional antennas and higher antenna beam widths perform well even if deployed at lower altitudes. Another challenge lies in determining the continuous UAV trajectory as it involves a large number of variables. A solution to this has been discussed in Ref. [54].
2.2.5 Trajectory Optimization
The performance of UAV assisted wireless networks can be significantly improved in aspect of throughput as well as coverage by optimizing the trajectory of the UAVs. This optimization depends upon the factors like flight constraints, energy constraints, ground user’s demands, collision avoidance, channel variations, mobility of UAV, etc. Table 2.2 lists the work carried out till date for optimizing the performance of the UAV systems by designing the optimum UAV trajectory.
The article by Chen et al. [62] proposes autonomous UAV wherein positions of the UAVs are self-optimized based on real time radio measurement.
Table 2.2 State-of-the-art solutions for optimizing the UAV trajectory.
| Parameter optimized | Effect on performance of system | Research article |
| User scheduling and trajectory of UAV | Maximized the minimum average data rate experienced by ground users | [55] |
| Trajectory of UAV with multiple antennas | Maximized system rate in uplink communication | [56] |
| Joint optimization of UAV trajectory and source/relay transmit power | Maximized throughput of relay based UAV system | [57] |
