No abstract

In this paper, a novel method for joint resource allocation, three-dimensional placement, and user association as a predominant problem in the internet of things (IoTs) networks is proposed. We consider two kinds of channel models for the links between users and aerial base stations (ABSs), namely (a) line of sight (LoS) model and (b) generalized model. In the LoS case, all the ABSs should establish LoS communications path toward IoT users. In the generalized case, the ABSs channels toward the IoT users could be LoS or nonline of sight. We propose four different scenarios, namely interference-free (IF)-LoS, IF-Generalized, interference-included (II)-LoS, and II-Generalized. The IF-LoS and IF-Generalized cases consider subcarrier and modulation order as the resources while the II-LoS and II-Generalized cases consider transmit power, subcarrier, and modulation order as the resources. Our objective function is minimizing the overall transmit power of the IoT users while satisfying some quality of service constraints in the uplink. To solve the resulting nonconvex optimization problem, we convert the main problem with high complexity into the subproblems with low complexities. We compare the performance of our joint proposed scenario with other joint and disjoint cases for the IF-LoS, IF-Generalized, II-LoS, and II-Generalized cases. Moreover, we evaluate the performance of the proposed schemes for different numbers of ABSs, environment types, target areas, and numbers of IoT users.Trans Emerging Tel Tech. 2019;30:e3632.wileyonlinelibrary.com/journal/ett

No abstract

In this paper, we propose a new scheme called dynamic power‐dynamic density at which our goal is to maximize the throughput of aerial base station–based networks with coverage probability and power density constraints. The corresponding optimization problem is nonconvex and intractable. To tackle this issue, we propose an iterative algorithm based on the well‐known alternative method. In this method, we decompose the optimization problem into two nonconvex subproblems, namely, power allocation and density finding. Then, in order to provide a low complex and high speed algorithm, these subproblems are transformed into geometric programming forms, which can be solved by existing convex tools. Moreover, the convergence of the proposed iterative scheme is proved and it is studied from the computational complexity perspective. To investigate the performance of the proposed algorithm with existing schemes, we consider two other schemes, namely, dynamic power‐fixed density and dynamic density‐fixed power. Based on simulation results, the throughput of aerial base stations in our scheme is approximately 27% more than that of the two existing schemes. Furthermore, the trade‐off between power and density for different network parameters is studied.