The exponential growth of the Internet of Things (IoT) has revolutionized the digital domain, allowing interconnected devices to share and process data with unprecedented efficiency [1,2]. Along with the growth in IoT devices, the demand for powerful and decentralized data processing solutions has led to the adoption of fog computing [3,4]. Unlike traditional cloud computing [5], it operates at the edge of a network [6], providing localized processing, storage, and control services to reduce latency and therefore improve real-time response [7,8]. This paradigm faces many challenges because, in general, fog nodes are resource-restricted compared to centralized cloud servers [9,10]. Most existing methods suffer from a lack of scalability [11], as the dense volume and diversity of IoT devices can rapidly overwhelm processing resources [12]. In addition, most models include machine learning and deep learning algorithms that are computationally intensive and thus may impede real-time responsiveness [13,14]. Energy efficiency is another major concern because most existing models do not balance processing needs with power limitations of edge devices [15]. With these challenges, the need for innovation in resource management techniques that can dynamically adapt to real-time demand is felt without the expense of unnecessary computational and energy overhead [16,17].

This work is motivated by the development of a framework for resource management to address the limitations found in existing approaches. Although useful, current methods often suffer from a general lack of flexibility and efficiency in providing scalable and sustainable IoT fog ecosystems [18]. This research aims to fill this gap by proposing a new quantum-inspired resource management algorithm [19], able to work within the limit of limited computing resources and energy availability. The central problem lies in achieving adaptive, energy-efficient resource allocation that aligns with the dynamic demands of real-time IoT applications [20]. Therefore, this study poses the following research question: How can a quantum-inspired adaptive resource management (QIARM) algorithm enhance scalability, energy efficiency, and adaptability in fog computing environments?

The proposed approach integrates principles from quantum computing, particularly quantum superposition, and entanglement for resource management in fog computing, as shown in Fig. 1. The proposed algorithm incorporates a quantum-inspired data structure that enables the simultaneous representation of multiple states of resource allocation. By simulating superposition, QIARM parallelly explores possible allocations and is thus much faster in its decisions. In addition, it proposes an entanglement-inspired mechanism to enable effective resource coordination, which simultaneously synchronizes the resource status of distributed fog nodes to mitigate redundant allocation and latency issues. In addition to this, QIARM includes an adaptive learning part that adjusts the strategies for allocation at runtime, according to network dynamics, and strikes a proper trade-off between performance and energy efficiency. Then, the integrated energy-aware scheduling model will ensure minimum consumption of power by favoring low-energy resource states with no sacrifice of computational effectiveness. The main contributions of this research study can be summarized as follows:

The proposed algorithm introduces a novel data structure inspired by quantum superposition that enables simultaneous consideration of multiple resource states.

The structure of this article can be summarized as follows: Section 2 reviews existing approaches to resource management to identify relevant research gaps. Section 3 introduces the Quantum Inspired Adaptive Resource Management (QIARM) approach, detailing its core components and methodology. Section 4 presents the experimental setup and simulation results to evaluate the QIARM performance. Section 6 provides a discussion of the results and their implications. Finally, Section 7 concludes the article.

The rapid proliferation of IoT devices has significantly increased the demand for resource-efficient computing and real-time data processing capabilities. Fog computing, which allows computation at the edge of the network, has emerged as a solution to address latency and bandwidth constraints in IoT environments [21,22]. Recent research has focused on developing efficient and adaptive resource management strategies to optimize performance and ensure responsiveness under varying conditions [23,24]. In fog computing environments, task offloading and resource allocation have become central to reducing delays and improving the quality of service for IoT applications [25,26]. Advanced resource management techniques have been proposed to address specific requirements, such as network splitting in 5G and 6G systems [27,28], retransmission strategies in the Long Range Wide Area Network (LoRaWAN) [29], and adaptive management of greenhouse environments [30].

Various approaches to resource management leverage machine learning and deep learning techniques to handle the complexities of IoT networks. For example, lightweight frameworks for distributed computing aim to balance computational load across IoT nodes, effectively managing resource constraints [31]. In addition to classifier systems, fog-cloud hybrid models have also been introduced to improve resource utilization by making intelligent decisions in distributed environments [32]. In Industrial IoT, dual-driven models have been presented based on reinforcement learning to improve sustainable computing [33]. Digital twin and blockchain-supported environments have also been suggested [34,35]. Furthermore, reinforcement learning has been used to enable resource management for deep neural network inference and improve computational efficiency in IoT networks where real-time analytics plays a key role [36].

Novel adaptive frameworks in IoT are still emerging that will include predictive maintenance and prioritization of resources to improve the reliability of the service and reduce any interruption to it [37,38]. This has been achieved in resource management on mobile edge computing using a priority mechanism to optimize resource allocation and achieve low latency for mobile users [26]. Pervasive edge computing models have also been popular, especially for applications that require very high data transmission rates, including VR streaming wirelessly in industrial IoT applications [39]. However, it has yet to be overcome concerning developing a scalable, adaptable, and energy-efficient resource management framework that can cope with the fluctuating demands of real-time IoT applications.

Another key limitation of existing resource management strategies is that they cannot respond appropriately to rapid changes in network load and availability, as reflected in predictive maintenance and priority-based models for mobile edge environments [26,37]. These models also do not consider any coordination mechanisms required by applications for sharing resources at different nodes, which is typical of latency-sensitive applications such as virtual reality streaming in industrial IoT [39]. The proposed QIARM approach aims to address challenges by introducing quantum-inspired resource representation and entanglement-like coordination mechanisms for runtime adjustments of resources. This allows QIARM to represent multiple states of allocation in superposition, leading to searching larger solution spaces while reducing computational overhead and deploying faster resource allocations. Furthermore, the energy-aware scheduling of QIARM optimizes energy consumption by adapting the allocation according to run-time conditions, which is not supported by state-of-the-art models. Fig. 2 summarizes the key limitations in existing methods and underlines how the proposed approach is addressing these challenges.

Quantum-inspired adaptive Resource Management switches on quite a new conceptual framework toward optimized resource allocation in fog computing environments. Featuring quantum-inspired superposition and entanglement, QIARM dynamically coordinates the distributed resources over time to make it adaptive with low latency and energy efficiency. This allows for addressing scalability challenges and enables real-time responsiveness uniquely for IoT applications over resource-constrained networks.

The simulation setup for the QIARM approach was executed on a high-performance computing platform that had an Intel Core i9 processor, 64 GB RAM, and an NVIDIA RTX 3090 GPU. We have used MATLAB R2022b as an environment to simulate fog and a multitude of IoT nodes under various demand conditions. The simulation parameters are designed to represent real-time IoT scenarios to the task offloading rate and energy consumption, considering the network latency. Table 1 summarizes the parameters used in our experiments, where we compare QIARM with the state-of-the-art approaches: Fog Resource-Based Adaptive Task Offloading (FRATO) [25], Adaptive Resource Management for Network Slice Architectures in 5G and 6G Systems (ADAPTIVE6G) [27], Retransmission-Assisted Resource Management (R-ARM) ** [29] and IoT-Pi [31]. This section presents an overview of the experimental environment, and Table 1 describes the simulation parameters to evaluate the performance of QIARM compared to the state-of-the-art models.

This section evaluates the computational complexity of the QIARM algorithm, focusing on its three core phases and comparing its overhead with baseline approaches.

The evaluation of the QIARM model shows notable superiority in resource management efficiency over some established models, such as FRATO, ADAPTIVE6G, R-ARM, and IoT-Pi. In all scenarios, QIARM can maintain a much higher task-offloading success rate along with much better resource utilization compared with other established schemes. Therefore, the performance adaptability and robustness of the QIARM are confirmed with fluctuating network conditions. While classical schemes fail in general for high-demand or variable-load conditions, the quantum-inspired algorithms from QIARM achieve sophisticated balancing of resources with minimal latency and energy consumption. Hence, the performance shows that quantum-inspired models can attempt to solve some major limitations in the existing resource management techniques under diverse scenarios and challenging IoT and fog computing conditions.

In addition, the adaptive resource allocation and scheduling mechanisms of QIARM helped achieve high energy efficiency, represented by the consumption of lower energy in all scenarios compared to baseline models. The quantum-based framework of QIARM dealt with real-time demands in a very nontraditional manner with an energy-aware perspective that made it rather different from the traditional models, which do not possess such responsive adaptability. These results show the feasibility of integrating quantum-inspired methods in the pursuit of greater efficiency in the IoT, especially in complex systems that are fully decentralized and require both resource flexibility and energy optimization. Although FRATO achieved a maximum task load success rate of 94% under low-demand conditions, our QIARM model achieved 98%, demonstrating superior adaptability. Additionally, QIARM reduced energy consumption by up to 20% across various scenarios compared to state-of-the-art methods like ADAPTIVE6G and IoT-Pi, which lacked similar energy-aware scheduling mechanisms. These results underline QIARM’s ability to address the scalability and efficiency challenges in fog computing environments, unlike traditional models that fail under dynamic network conditions.

Although QIARM implements effective optimization in resource allocation, quantum-inspired mechanisms may introduce computational overheads that could affect scalability for larger networks. Future research efforts should be made to further refine and develop these algorithms to reduce computational overheads so that, in the future, quantum-inspired methods remain viable even for large-scale networks in the IoT. Furthermore, hybrid solutions that merge quantum-inspired principles with other emerging models can also be further explored to achieve even higher improvements in performance and energy efficiency.

Efficient resource management in IoT and fog computing environments is increasingly critical to meet the demands of scalable and energy-efficient networks. This study introduced the QIARM model, designed to optimize resource allocation in dynamic conditions. Unlike conventional methods, QIARM uniquely leverages quantum principles to balance resource use and energy efficiency through adaptive scheduling. Practically, this innovation has potential applications in diverse IoT networks, where flexible resource management is essential. The evaluation showed QIARM’s superior performance, with a task-offloading success rate reaching 98% under low-demand conditions, compared to 94% for FRATO and 92% for ADAPTIVE6G. Under high-demand scenarios, QIARM maintained a utilization rate of 89%, outperforming FRATO’s 82% and R-ARM’s 77%. Furthermore, QIARM demonstrated notable energy efficiency, reducing consumption by up to 20% in all scenarios. Although the QIARM model demonstrates significant improvements in scalability, energy efficiency, and resource management, it introduces computational overhead due to the quantum-inspired mechanisms. Future extensions could explore hybrid models, integrating quantum-inspired principles with machine learning for even greater optimization in complex IoT environments.