One of the primary requirements in cellular-enabled multiunmanned aerial vehicle (UAV) Internet of Things (IoT) networks is to preserve data rates according to the IoT users’ (IUs) requirements. The mobility of IUs, 3-D movement of UAVs, environmental conditions, and bandwidth allocation to the IUs increase the challenges to maintain the data rates due to the continual change in the channel state. A constant extraction of channel state is crucial in this regard. We construct a sum-rate maximization problem considering the channel-state information (CSI). Unlike previous work, we propose CSI-driven data rate optimization for multi-UAV IoT networks (CARTEL). First, it allocates optimized bandwidth to IUs and accomplishes UAV–IU associations by adopting the matrix minima method. Subsequently, it maximizes the sum-rate invoking four modules: 1) parameter selector (PS); 2) IU tracker (IT); 3) path-loss estimator (PE); and 4) policy generator (PG). PS, IT, and PE help to extract the CSI by selecting suitable environmental parameters, tracking the IU mobility, and estimating the path loss, respectively. Finally, PG maximizes the data rates by adjusting the 3-D position and transmitting the power of a UAV. Extensive simulation results depict that the sum-rate in CARTEL improves by 26.03% and 65.46% than learn-as-you-fly (LAYF) and random selection (RS), respectively.

One of the crucial challenges in networked Unmanned Aerial Vehicles (UAVs) is to configure them to serve as aerial base stations (BSs) for collecting data from distributed Internet of Things (IoT) devices in a region devoid of backbone connectivity. To address this challenge, it is required to compute optimized trajectories of UAVs to collect data while considering the different activation patterns of IoT devices. We propose a scheme to optimize the trade-off between the number of covered IoT devices and travel time of UAVs. The formulated cost minimization problem is known as the capacitated single depot vehicle routing problem (CSDVRP), which is NP-hard. We propose a solution scheme, named Cost-Effective Data Aggregation for UAV-Enabled IoT Networks (CEDAN), which operates in four steps. First, it determines the optimized hovering locations (HLs) for UAVs. Subsequently, CEDAN determines the optimized route adopting the Christofides's approximation algorithm for Travelling Salesman Problem (TSP). Further, a split function produces the optimized trajectories for all UAVs. Finally, a route adjustment algorithm applies the cost function and rearranges the order of visiting each HL. Extensive simulation results depict that the CEDAN outperforms than Clarke-Wright (CW) savings heuristics, CEDAN without route adjustment (CWRA), and Zhan et al., respectively.

This work presents an efficient data-centric client selection approach, named DICE, to enable federated learning (FL) over distributed edge networks. Prior research focused on assessing the computation and communication ability of the client devices for selection in FL. On-device data quality, in terms of data volume and heterogeneity, across these distributed devices is largely overlooked. The obvious outcome is the selection of an improper subset of clients with poor-quallity data, which inevitably results in an inefficient trained model. With an aim to address this problem, in this work, we design DICE which prioritizes the data quality of the client devices in the selection phase, in addition to their computation and communication abilities, to improve the accuracy of FL. Additionally, in DICE, we introduce the assistance of vicinal edge devices to account for the lack of computation or communication abilities in certain devices without violating the privacy-preserving guarantees of FL. Towards this aim, we propose a scheme to decide the optimal edge device, in terms of latency and workload, to be selected as the helper device. The experimental results show that DICE improves convergence speed for a given level of model accuracy. Further, the simulation results show that DICE reduces delay by at least 16%, energy consumption by at least 17%, and packet loss by at least 55% compared to the existing benchmarks while prioritizing the on-device data quality across clients.

This work sketches a location priority-aware service availability scheme for use in a cellular-enabled multiple unmanned aerial vehicle (UAV) networks for Internet-of-Things (IoT) applications. These UAVs have the ability to sense the location-based data employing onboard heterogeneous sensors and send the sensed IoT data to the base station (BS). Existing works treat all locations equally and are inefficient in terms of assigning UAVs to provide on-demand location-based IoT services to the users. To address this issue, we formulate the objective as a optimization problem to maximize service availability. As the formulated objective is NP-hard, we propose a scheme named priority-aware service availability for multi-UAV networks (PRISM) to obtain the solution. PRISM operates in two steps. First, it offers a UAV assignment algorithm to prioritize locations and assign the UAVs to the high-priority locations proactively. Next, it employs a service assignment algorithm that assigns the requested location-based IoT services to UAVs. The results of simulation and real experiments depict the efficacy of PRISM in terms of availability and delay of the incoming IoT service requests. From the results, we observe that PRISM improves the average service availability by 20.61% and reduces average service delay by 29.4%, compared to the benchmark solutions.