As intelligent network services continue to diversify, ensuring efficient and adaptive resource allocation in edge networks has become increasingly critical. Yet the wide functional variations across services often give rise to new and unforeseen optimization problems, rendering traditional manual modeling and solver design both time-consuming and inflexible. This limitation reveals a key gap between current methods and human solving — the inability to recognize and understand problem characteristics. It raises the question of whether problem-aware learning can bridge this gap and support effective cross-problem generalization. To answer this question, we propose a problem-aware diffusion (PAD) model, which leverages a problem-aware learning framework to enable cross-problem generalization. By explicitly encoding the mathematical formulations of optimization problems into token-level embeddings, PAD empowers the model to understand and adapt to problem structures. Extensive experiments across ten representative network optimization problems show that PAD generalizes well to unseen problems while avoiding the inefficiency of building new solvers from scratch, yet still delivering competitive solution quality. Meanwhile, an auxiliary constraint-aware module is designed to enforce solution validity further. The experiments indicate that problem-aware learning opens a promising direction toward general-purpose solvers for intelligent network operation and resource management. Our code is open source at https://github.com/qiyu3816/PAD.

Network optimization is a fundamental challenge in the Internet of Things (IoT) network, often characterized by complex features that make it difficult to solve these problems. Recently, generative diffusion models (GDMs) have emerged as a promising new approach to network optimization, with the potential to directly address these optimization problems. However, the application of GDMs in this field is still in its early stages, and there is a noticeable lack of theoretical research and empirical findings. In this study, we first explore the intrinsic characteristics of generative models. Next, we provide a concise theoretical proof and intuitive demonstration of the advantages of generative models over discriminative models in network optimization. Based on this exploration, we implement GDMs as optimizers aimed at learning high-quality solution distributions for given inputs, sampling from these distributions during inference to approximate or achieve optimal solutions. Specifically, we utilize denoising diffusion probabilistic models (DDPMs) and employ a classifier-free guidance mechanism to manage conditional guidance based on input parameters. We conduct extensive experiments across three challenging network optimization problems. By investigating various model configurations and the principles of GDMs as optimizers, we demonstrate the ability to overcome prediction errors and validate the convergence of generated solutions to optimal solutions. We provide code and data athttps://github.com/qiyu3816/DiffSG.

Optimization is crucial for the efficiency and reliability of multi-access edge computing (MEC) networks. Many optimization problems in this field are NP-hard and do not have effective approximation algorithms. Consequently, there is often a lack of optimal (ground-truth) data, which limits the effectiveness of traditional deep learning approaches. Most existing learning-based methods require a large amount of optimal data and do not leverage the potential advantages of using suboptimal data, which can be obtained more efficiently. To illustrate this point, we focus on the multi-server multi-user computation offloading (MSCO) problem, a common issue in MEC networks that lacks efficient optimal solution methods. In this paper, we introduce the graph diffusion-based solution generator (GDSG), designed to work with suboptimal datasets while still achieving convergence to the optimal solution with high probability. We reformulate the network optimization challenge as a distribution-learning problem and provide a clear explanation of how to learn from suboptimal training datasets. We develop GDSG, a multi-task diffusion generative model that employs a graph neural network (GNN) to capture the distribution of high-quality solutions. Our approach includes a straightforward and efficient heuristic method to generate a sufficient amount of training data composed entirely of suboptimal solutions. In our implementation, we enhance the GNN architecture to achieve better generalization. Moreover, the proposed GDSG can achieve nearly 100% task orthogonality, which helps prevent negative interference between the discrete and continuous solution generation training objectives. We demonstrate that this orthogonality arises from the diffusion-related training loss in GDSG, rather than from the GNN architecture itself. Finally, our experiments show that the proposed GDSG outperforms other benchmark methods on both optimal and suboptimal training datasets. Regarding the minimization of computation offloading costs, GDSG achieves savings of up to 56.62% on the ground-truth training set and 41.06% on the suboptimal training set compared to existing discriminative methods.

Computation offloading has become a popular solution to support computationally intensive and latency-sensitive applications by transferring computing tasks to mobile edge servers (MESs) for execution, which is known as mobile/multi-access edge computing (MEC). To improve the MEC performance, it is required to design an optimal offloading strategy that includes offloading decision (i.e., whether offloading or not) and computational resource allocation of MEC. The design can be formulated as a mixed-integer nonlinear programming (MINLP) problem, which is generally NP-hard and its effective solution can be obtained by performing online inference through a well-trained deep neural network (DNN) model. However, when the system environments change dynamically, the DNN model may lose efficacy due to the drift of input parameters, thereby decreasing the generalization ability of the DNN model. To address this unique challenge, in this paper, we propose a multi-head ensemble multi-task learning (MEMTL) approach with a shared backbone and multiple prediction heads (PHs). Specifically, the shared backbone will be invariant during the PHs training and the inferred results will be ensembled, thereby significantly reducing the required training overhead and improving the inference performance. As a result, the joint optimization problem for offloading decision and resource allocation can be efficiently solved even in a time-varying wireless environment. Experimental results show that the proposed MEMTL outperforms benchmark methods in both the inference accuracy and mean square error without requiring additional training data.