With the widespread application of the Internet of Things (IoT), computing tasks on the terminal side have surged. Traditional cloud computing models, constrained by high network latency and overloaded central servers, can no longer effectively meet the dual requirements of real-time responsiveness and energy efficiency. The cloud–edge–device collaborative architecture, by enabling distributed resource scheduling, offers a promising solution to reduce both latency and energy consumption. However, optimizing carbon emissions under dynamic operating conditions remains a pressing and unresolved challenge. This paper proposes a carbon-aware dynamic scheduling framework for cloud–edge–device systems, which accounts for the stochastic nature of task arrivals, heterogeneous computing capabilities, and varying carbon intensity across devices and locations. A multi-layer carbon emission model is developed, and the long-term carbon minimization objective is formulated as a stochastic optimization problem. Using the Lyapunov drift-plus-penalty method, the problem is transformed into a tractable deterministic optimization framework, upon which a Carbon-Efficient Computation Offloading (CECO) algorithm is designed. CECO jointly optimizes local computation frequency, data transmission rate, and edge resource allocation to dynamically balance task queue stability and carbon emission intensity. Theoretical analysis and simulation results validate that the proposed algorithm significantly reduces system-level carbon emissions while maintaining quality of service, demonstrating strong potential for enabling green computing in intelligent distributed environments.

ABSTRACT The exponential growth of the Internet of Things (IoT) and Internet of Medical Things (IoMT) has imposed significant demands on cloud‐centric architectures, particularly in terms of latency, energy efficiency, and scalability. Fog computing emerges as a promising paradigm by decentralizing computation and storage closer to data sources. This paper presents and validates a unified framework that integrates multi‐objective metaheuristic optimization for computation offloading and task scheduling with graph‐based database architectures for real‐time IoMT data management. The key novelty lies in the system‐level integration of these components, creating a feedback‐aware architecture where scheduling decisions explicitly account for database performance, rather than proposing new optimization algorithms in isolation. The study leverages enhanced metaheuristic algorithms such as Multi‐objective Arithmetic Optimization Algorithm (MoAOA) and Henon‐Evoked Rhinopithecus Swarm Optimization (HERSOA) to optimize energy consumption, latency, and throughput while ensuring reliable task prioritization and resource allocation. A formal data‐layer latency model is introduced and validated, with the optimization framework explicitly incorporating database performance metrics. Through comprehensive simulation, ablation studies, and statistical validation, we demonstrate that the combination of advanced optimization techniques and graph‐based fog architectures significantly improves QoS parameters, system stability, and scalability in dynamic IoT environments. This work provides a foundational guideline for future research and practical deployments in smart healthcare, smart cities, and industrial IoT.

With the development of intelligent transportation systems, vehicular applications demonstrate diverse characteristics, including computation-intensive processing and stringent latency requirements. Traditional computation offloading strategies struggle to cope with the highly dynamic, multi-node, and multi-task concurrent vehicular network environment and generally overlook the risk of cross-zone communication failures caused by high-speed mobility. To address this issue, this paper designs a computation offloading algorithm based on multi-agent reinforcement learning. This method comprehensively considers four heterogeneous features including queue load, communication links, task attributes, and computing resources, establishes a multi-layer collaborative computing architecture integrating task migration and result return mechanisms, and further constructs an optimization model aimed at minimizing the weighted sum of latency and energy consumption. This model is formalized as a multi-agent Markov decision process, and an improved Multi-Agent Proximal Policy Optimization(MAPPO)-based MATPPO-T algorithm is designed to solve it, achieving one-step joint optimization of task offloading, resource allocation, and task result migration. Experimental results demonstrate that the proposed method reduces the total system cost by approximately 22% on average compared to benchmark algorithms such as MAPPO and PPO, while consistently maintaining the lowest offloading overhead and fastest convergence speed, validating its robustness and scalability in dynamic vehicular edge networks.

Computation offloading and scheduling in mobile edge networks and SDN

The rapid development of space exploration demands real-time backhaul of massive sensing payload data in space-ground integrated telemetry, tracking, and command (TT&C) networks. However, traditional narrow-band TT&C links suffer from severe congestion during massive data backhaul. Since most TT&C applications are inherently task-oriented and do not require pixel-perfect data reconstruction, we propose a task-oriented joint resource allocation framework based on semantic communications. Specifically, we introduce an adaptive semantic split computing mechanism that extracts and transmits only compact, decision-critical features instead of raw bitstreams, fundamentally mitigating the bandwidth bottleneck. The joint optimization of computation offloading, semantic splitting, and continuous on-board computing allocation is formulated as a stochastic mixed-integer nonlinear programming (MINLP) problem. We propose a decoupled algorithm based on Hierarchical Multi-Agent Proximal Policy Optimization (HMAPPO) to solve it. An outer layer employs multi-agent reinforcement learning (MARL) for distributed discrete decision-making, while an inner layer utilizes a Karush–Kuhn–Tucker (KKT)-based solver for continuous space-based computing allocation. This bi-level architecture overcomes the curse of dimensionality and mathematically guarantees zero-violation of physical capacity constraints. Simulations demonstrate that HMAPPO rapidly converges and sustains a high weighted success rate under heavy traffic congestion, significantly improving system utility compared to state-of-the-art baselines.

To address the issue of Quality of Service (QoS) degradation for mission-critical tasks in Low Earth Orbit (LEO) satellite edge computing under extreme congestion, this paper proposes an intelligent computation offloading architecture based on graph attention imitation learning. The proposed method models the satellite network as a time-varying graph to extract global topological features and innovatively introduces a QoS-aware action masking mechanism to forcibly preclude suboptimal decisions through strict physical constraints. Simulation results demonstrate that the proposed algorithm significantly reduces both the network-wide average latency and the 95th percentile tail latency. Furthermore, it achieves the highest success rate for high-priority tasks under heavy-load conditions. Ultimately, this architecture effectively overcomes the "herd effect" among local nodes, realizing superior global load balancing and absolute QoS guarantees for critical tasks within dynamic satellite networks.

Towards dynamic pricing for computation offloading in mobile edge computing: a federated learning approach

The unmanned aerial vehicle (UAV)-assisted space-air-ground integrated networks (SAGIN) can provide communication and computing services for Internet of Remote Things (IoRT) in the absence of ground cellular network coverage. In this paper, we propose an edge computing architecture based on SAGIN, comprising three parts: a satellite, UAVs, and ground-based IoRT devices. The satellite is responsible for providing access to cloud computing resources. UAVs are equipped with mobile edge computing (MEC) servers. And IoRT devices generate latency-sensitive tasks but possess limited computing capabilities. Our objective is to minimize task processing delays by jointly optimizing UAV deployment, computation offloading, and time slot resource allocation to meet the increasing demands of IoRT devices. Specifically, the proposed problem is decomposed into two components. First, for the deployment of multiple UAVs, we propose a multi-agent softmax deep double deterministic policy gradient (MASD3) approach, which enables UAVs to adjust their flight trajectories based solely on observed information for adaptive deployment. Second, for the computation offloading and time slot resource allocation problems in SAGIN, we employ a numerical computation-based iterative optimization method to minimize the occupation of time slots by computation offloading. Simulation results demonstrate that our proposed solution significantly reduces overall delay compared to alternative benchmark schemes.

Mobile edge computing (MEC) allows ground vehicles (GVs) to offload computationally intensive tasks to edge servers, offering advantages over centralized cloud computing by reducing the energy consumption and network congestion. However, the dynamic nature of communication links often results in suboptimal offloading performance due to signal occlusion and interference. Reconfigurable intelligent surface (RIS) technology, a promising component of sixth-generation (6G) communication networks, can enhance wireless network capabilities by modifying the phase and amplitude of reflective components. In this paper, we propose a RIS-assisted MEC strategy to provide a distributed edge intelligence (DEI) solution for systems providing fast wireless connectivity and low latency to ground vehicles in dynamic environments. The RIS-assisted vehicular networks model has recently showed promising results when the delay was minimized by jointly optimizing the computation offloading strategy and the RIS phase shift. The delay optimization problem was modeled as a markov decision process (MDP), and a delay minimization algorithm (DDPG-DM) based on deep reinforcement learning (DRL) was proposed. Simulation results demonstrate that the proposed algorithm significantly outperforms existing non-RIS learning algorithms and classical methods, achieving superior performance in reducing delay. The findings suggest that integrating RIS with MEC can substantially improve the efficiency of computation offloading in dynamic vehicular environments.

Unmanned Aerial Vehicle (UAV) computing clusters face severe operational constraints due to limited computing capabilities and battery capacities, which complicate the simultaneous optimization of low offloading latency, long task endurance, and high cluster efficiency. To address these challenges, this paper proposes a Multi-Objective Reinforcement Learning framework based on Latency and Power Balance (MORL-LAPB). Instead of broad situational awareness descriptions, our framework directly combines a reward-shaping reinforcement learning algorithm with an evolutionary mechanism to construct a closed-loop optimization paradigm. Crucially, in this context, 'balancing' extends beyond traditional computational workload distribution; it represents a joint optimization that balances task allocation to ensure short service delays while simultaneously equating the energy depletion rates across UAV nodes to maximize overall cluster efficiency and operational duration. By efficiently identifying Pareto optimal trade-offs, MORL-LAPB dynamically regulates UAV energy allocation and computational resource scheduling. Experimental results demonstrate that, compared to RSO, NSO, and DRLSO baselines, the proposed MORL-LAPB significantly reduces offloading latency, extends effective task execution duration, and improves cluster energy efficiency. The framework offers flexible adaptability and long-term sustainability for diverse operational scenarios under strict multi-objective constraints.

SGICPNOM: A computation offloading mechanism for 6G space-ground integrated computing power network

The unmanned aerial vehicle (UAV) based multi-access edge computing (MEC) appears as a popular paradigm to reduce task processing latency. However, the secure offloading is an important issue when occurring aerial eavesdropping. Besides, the potential uncertainties in practical applications and flexible trajectory optimizations of UAVs pose formidable challenges for realizing robust offloading. In this paper, we consider the aerial secure MEC network including ground users, service unmanned aerial vehicles (S-UAVs) integrated with edge servers, and malicious UAVs overhearing transmission links. To deal with the task computation complexities, which are characterized as uncertainties, a robust problem is formulated with chance constraints. The energy cost is minimized by optimizing the connections, trajectories of S-UAVs and offloading ratios. Then, the proposed non-linear problem is tackled via the distributionally robust optimization and conditional value-at-risk mechanism, which is further transformed into the second order cone programming forms. Moreover, we decouple the reformulated problem and design the successive convex approximation for S-UAV trajectories. The global algorithm is designed to solve the sub-problems in a block coordinate decent manner. Finally, extensive simulations and numerical analyses are conducted to verify the robustness of the proposed algorithms, with just 2% more energy cost compared with the ideal circumstance.

Adaptive task offloading and resource orchestration under spatiotemporal constraints in mobile edge computing

Deep reinforcement learning-based computation offloading and resource allocation in user-centered UAV-MEC

Nevertheless, mobile devices harness resource poverty, which includes limited processing power, limited battery life, and limited memory storage, and this is still a considerable drawback to the provision of intensive applications, including 3D simulations and data analytics in real-time. The paper will suggest a strong three-tier hybrid architecture providing integration of mobile computing and cloud services to provide seamless delivery of education. The architecture consisted of a User Interface Tier of lightweight front-end interactions, a Middleware Tier of a Task Orchestrator and a Threshold-Based Elastic Scaling (TBES) algorithm, and a Data Tier of hybrid NoSQL storage for real-time synchronization. Created a mathematical model of energy-aware offloading to estimate the optimal point of offloading, depending on network latency and the CPU intensity locally. Simulation outcomes show that there are great performance increases in comparison with the traditional local-execution models. The TBES algorithm was able to keep the system response time at 185 ms, which is much lower than the 200 ms Rmax that is required, and even at the maximum verified load of 500 users. Furthermore, computation offloading extended mobile battery longevity by 73.3%, reducing power consumption from a 450-mW baseline to 120 mW. Bandwidth efficiency was also optimized through a differential synchronization protocol, which achieved an 88.5% reduction in data overhead. Correlation analysis using Kaggle education datasets further suggests that these technical enhancements contributed to a 22% increase in course completion rates and a 92% improvement in student engagement. These results confirm that the suggested architecture can efficiently isolate the quality of education on the one hand and the limitations of local hardware on the other, providing a solution that can be scaled to comprehensive, global e-learning.

Deep learning approaches for computation offloading in edge computing: A critical review

Abstract Secure computation offloading is challenging because sensitive data can be exposed to malicious users, especially when the risk of data leakage arises during transmission or processing in untrusted edge environments. Therefore, in this paper, we present a novel content-aware data leakage prevention framework that classifies computational tasks into low, medium, and high sensitivity levels. We propose to address the trade-off between energy consumption and privacy under time constraints by integrating digital twin technology for dynamic and real-time network control. Compared with traditional edge computing and IoT environments, digital twin networks provide a virtualized, continuously synchronized representation of physical entities and enables more precise, context-aware decision-making for task offloading and security management. This bidirectional mapping between the physical and virtual worlds allows for proactive risk detection, adaptive privacy control, and optimized resource allocation, which are not achievable in conventional edge or IoT systems. In this regard, we utilize the digital twin’s real-time, synchronized representation of physical entities to design an energy and privacy-aware task scheduling scheme and minimize data leakage risks in digital twin edge networks. To achieve these, we analytically model both energy consumption and task processing delay using the M/M/1 queuing model. Furthermore, we propose an energy and privacy-aware genetic algorithm-based task scheduling algorithm designed to minimize energy consumption and latency while maximizing privacy in digital twin edge networks. Under strict latency constraints, numerical results show that while the conventional approach can handle up to 400 tasks, the proposed model scales efficiently to 600 tasks by prioritizing energy efficiency and privacy, and improves scalability by 1.5 times. Here, increasing the number of iterations has a much more positive effect than increasing the number of chromosomes. It achieves at most 84% privacy gain and acceptable 0.197 watts energy consumption in the topology with 0.58 second response time. In addition, the proposed approach offers the ability to further improve processing time and privacy gains through digital twin monitoring.

Unmanned Aerial Vehicle (UAV)-assisted edge computing supports latency-sensitive applications by offloading computational tasks to ground-based servers. However, determining optimal resource allocation under strict latency constraints and stochastic channel conditions remains challenging. This paper addresses the joint computation partitioning and power allocation problem for UAV-assisted edge computing systems. We formulate the problem as a Constrained Markov Decision Process (CMDP) that explicitly models latency constraints, rather than relying on implicit reward shaping. To solve this CMDP, we propose Constrained Soft Actor-Critic (C-SAC), a deep reinforcement learning algorithm that combines maximum-entropy policy optimization with Lagrangian dual methods. C-SAC employs a dedicated constraint critic network to estimate long-term constraint violations and an adaptive Lagrange multiplier that automatically balances energy efficiency against latency satisfaction without manual tuning. Extensive experiments demonstrate that C-SAC achieves an 18.9% constraint violation rate. This represents a 60.6-percentage-point improvement compared to unconstrained Soft Actor-Critic, with 79.5%, and a 22.4-percentage-point improvement over deterministic TD3-Lagrangian, achieving 41.3%. The learned policies exhibit strong channel-adaptive behavior with a correlation coefficient of -0.894 between the local computation ratio and channel quality, despite the absence of explicit channel modeling in the reward function. Ablation studies confirm that both adaptive mechanisms are essential, while sensitivity analyses show that C-SAC maintains robust performance with violation rates varying by less than 2 percentage points even as channel variability triples. These results establish constrained reinforcement learning as an effective approach for reliable UAV edge computing under stringent quality-of-service requirements.

NOTICE OF RETRACTION FOR: Abdul-Samad, Sarmad T.; Al-Hwaidi, Osamah; and Muslem, Ali Abd Al-Rasool (2025) ``Efficient Multi-User Computation Offloading and Reducing Latency in Mobile-Edge Computing for IoT Applications,'' Iraqi Journal for Computer Science and Mathematics: Vol. 6: Iss. 3, Article 29. DOI: https://doi.org/10.52866/2788-7421.1298. Available at: https://ijcsm.researchcommons.org/ijcsm/vol6/iss3/29.

ABSTRACT Integrating blockchain technology with Internet of Things (IoT) networks presents opportunities and challenges for sustainable computing. While blockchain ensures secure and transparent data management, its energy‐intensive nature poses significant environmental concerns, particularly in resource‐constrained IoT environments. This paper proposes SERO‐DRL, a novel deep reinforcement learning approach for energy‐efficient resource optimization in blockchain‐enabled sustainable IoT networks. We develop a comprehensive framework that jointly optimizes computational offloading and resource allocation while considering renewable energy availability and environmental impact. The framework includes an innovative reward mechanism that incentivizes energy‐efficient behavior while ensuring fair resource allocation among IoT devices. Experimental results demonstrate SERO‐DRL's superior performance, achieving an 18.5% reduction in total system costs and a 40% decrease in environmental impact compared to baseline approaches.