Efficient optimization techniques for resource allocation in UAVs mission framework

PLOS ONE RESEARCH ARTICLE Efficient optimization techniques for resource allocation in UAVs mission framework Sohail Razzaq1,2, Costas Xydeas3, Anzar Mahmood4, Saeed Ahmed4, Naeem Iqbal Ratyal4,5, Jamshed Iqbal ID6* 1 Faculty of Information Technology, Majan University College, Muscat, Sultanate of Oman, 2 Department of Electrical and Computer Engineering, COMSATS University Islamabad, Abbottabad Campus, Abbottabad, Pakistan, 3 School of Computing and Communications, Lancaster University, Lancaster, United Kingdom, 4 Department of Electrical Engineering, Mirpur University of Science and Technology, Mirpur, Pakistan, 5 Department of Computer System Engineering, Mirpur University of Science and Technology, Mirpur, Pakistan, 6 School of Computer Science, Faculty of Science and Engineering, University of Hull, Hull, United Kingdom * a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 OPEN ACCESS Citation: Razzaq S, Xydeas C, Mahmood A, Ahmed S, Ratyal NI, Iqbal J (2023) Efficient optimization techniques for resource allocation in UAVs mission framework. PLoS ONE 18(4): e0283923. https:// doi.org/10.1371/journal.pone.0283923 Editor: Chakchai So-In, Khon Kaen University, THAILAND Received: July 29, 2021 Accepted: March 21, 2023 Published: April 6, 2023 Copyright: © 2023 Razzaq et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Data Availability Statement: All relevant data are within the paper. Funding: The author(s) received no specific funding for this work. Competing interests: The authors have declared that no competing interests exist. Abstract This paper considers the generic problem of a central authority selecting an appropriate subset of operators in order to perform a process (i.e. mission or task) in an optimized manner. The subset is selected from a given and usually large set of ‘n’ candidate operators, with each operator having a certain resource availability and capability. This general mission performance optimization problem is considered in terms of Unmanned Aerial Vehicles (UAVs) acting as firefighting operators in a fire extinguishing mission and from a deterministic and a stochastic algorithmic point of view. Thus the applicability and performance of certain computationally efficient stochastic multistage optimization schemes is examined and compared to that produced by corresponding deterministic schemes. The simulation results show acceptable accuracy as well as useful computational efficiency of the proposed schemes when applied to the time critical resource allocation optimization problem. Distinguishing features of this work include development of a comprehensive UAV firefighting mission framework, development of deterministic as well as stochastic resource allocation optimization techniques for the mission and development of time-efficient search schemes. The work presented here is also useful for other UAV applications such as health care, surveillance and security operations as well as for other areas involving resource allocation such as wireless communications and smart grid. 1. Introduction This paper considers the generic problem of selecting an appropriate subset of operators in order to perform a process (mission/task). The subset is taken from a given and usually large ‘n’ elements set, with each operator having a certain resource availability/capability. The aim, underpinning this selection procedure, is the successful and efficient completion of the mission. This is essentially a discrete optimization problem which requires knowledge of: i) a PLOS ONE | https://doi.org/10.1371/journal.pone.0283923 April 6, 2023 1 / 19 PLOS ONE UAV optimization corresponding mission/task model (MM), ii) a mission performance objective measure (MPOM) and iii) the resources available to operators and also associated resource related constraints. Furthermore in order to study such a generic problem, in conjunction with a real application scenario, a firefighting mission paradigm has been adopted. In this case the above optimization problem is addressed in terms of the deployment of Unmanned Aerial Vehicles (UAVs) acting as firefighting operators in a fire extinguishing mission. UAVs which can be deployed from a single or from multiple geographical positions, are therefore required to deliver fire extinguishing material(s) at the appropriate fire incident location (FL). Thus the selection of a subset of available UAVs in a way that maximizes mission benefit (performance), given certain resource related constraints, is effectively the nonlinear, binary variables optimization problem whose efficient solution is the underlining theme of the paper. The binary variables resource allocation optimization problem outlined above can be stated as: Minimize FðXÞ Subject to : AX � Q X 2 ð0; 1Þ ð1Þ n where F is a real valued, continuous and non differentiable function, X is the vector whose binary valued elements represent the absence (i.e. = 0) or participation (i.e. = 1) in the mission for each of the n available operators, A is a resource related matrix and Q is a vector representing resource constraints. In general, the problem of Resource Allocation (RA) can be addressed in a number of different ways. These may differ in the presentation of input data, the handling of constraints, the search method used for finding the best solution and the computational complexity and robustness of the optimization approach. There are also hybrid instances available in the literature of scenario-based presentations of input data. Mulvey, Vanderbei and Zenios [1], for example, produced an approach that considers optimization formulations with a scenariobased description of input data. Each scenario is associated with a set of possible instances of uncertain problem data and with the probability of occurrence of that scenario. The authors use this approach to enhance optimization robustness. Furthermore, they devised a model to measure the trade-off between algorithmic optimization quality and robustness. In [2], Kouvelis and Yu present a more detailed account of scenario-based robust optimization techniques. Deterministic optimization methods consider all the input variables and the resulting mission benefit as deterministic quantities whereas in stochastic optimization techniques, some of the input variables and hence the mission benefit are subject to uncertainty. In the second case, optimization robustness is of fundamental importance and is sought against variable uncertainty and worst case scenarios. A detailed account of deterministic optimization techniques is presented in [3]]. Note that variable and mission benefit uncertainty is a common feature of real applications where the use of resource allocation procedures is required. Various methods of implementing st (...truncated)