Caching contents in edge networks can reduce latency and lighten the burden on backhaul links. Since the capacity of cache nodes is limited, accurate content popularity distribution is crucial to the effectual usage of cache capacity. However, existing popularity prediction models stem from big data and, hence, may suffer poor accuracy due to the small population in edge caching. In this letter, we propose a social-driven propagation dynamics-based prediction model, which requires neither training phases nor prior knowledge. Specifically, we first explore social relationships to bridge the gap between small population and prediction accuracy under susceptible-infected-recovery model. Then, a discrete-time markov chain approach is proposed to predict the viewing probability of certain contents from the perspective of individuals. Simulations validate that our proposed model outperforms other solutions significantly, by improving up to 94% in accuracy and 99% less runtime overhead.

Edge computing allows an edge server to adaptively place virtual instances to serve different types of data. This paper presents a new algorithm which jointly optimizes virtual service placement farsightedly and service data admission instantly to maximize the time-average service throughput of edge computing. The data admission is optimized, adapting to fast-changing data arrivals and wireless channels. The service placement is transformed into a two-dimensional knapsack problem by approximating future arrivals and channels with past observations, and solved over a slow timescale to allow services to be properly installed. Different from existing studies, our algorithm considers practical aspects of edge servers, such as finite memory size and bandwidth. We prove that the algorithm is asymptotically optimal and the optimality loss resulting from the approximation diminishes. Simulations show that our approach can improve the time-average throughput of existing alternatives by 16% for our considered simulation setup. The improvement becomes higher, as the memory size becomes increasingly tight. The number of services to be replaced is reduced without loss of throughput, after being placed farsightedly.

The small cells have been equipped with caching and multicast capabilities to save energy and ease backhaul burden. However, given the increasing diversity of user association in heterogeneous networks, traditional schemes may fail to exploit the energy-saving potential of caching and multicast. In this paper, we propose the model to minimize total power consumption by jointly optimizing the user association and cache deployment. The formulated joint optimization problem is decomposed and reformulated as a NP-complete set partition problem. Motivated by the idea of the tactile networks, the devices can find the candidate user grouping based on its own utility. The utility function is judiciously designed to reduce the searching space without compromising optimality. Then, a heuristic caching algorithm is proposed by rigorously deriving the upper and lower bounds of cache placement. Simulation results show that our proposed scheme outperforms the other existing multicast and caching algorithms in terms of power consumption by up to 28%, while keeping the load among base stations balanced.

The rapid growth of e-commerce has intensified the demand for efficient urban logistics. Electric Vehicles (EVs), with their eco-friendly and high-efficiency features, have emerged as a promising solution for improving urban logistics efficiency. However, due to their limited battery capacity, EVs often require recharging during operations, and improper charging decisions may lead to delivery delays, resulting in a loss of platform revenue. In this paper, we explore a novel EV Charging-Aware Task Assignment (ECTA) problem in urban logistics scenarios, where the objective is to maximize platform revenue by ensuring timely task completion while meeting the charging needs of EVs. To address this challenge, we present e-Charge, an efficient two-stage framework that enables real-time optimization of two continuous processes: task assignment and charging decision. For task assignment, which focuses on matching tasks to suitable EVs, we construct a hybrid weight model that incorporates charging penalties to calculate matching weights for EVs in both active and charging states, thus improving task assignment quality. Additionally, we implement an effective vehicle selection strategy to expedite the matching process, ensuring the efficiency of task assignment. For charging decision, which focuses on determining when and where EVs should be charged, we propose a multi-agent reinforcement learning (MARL) approach to dynamically select the charging timing for EVs. To further enhance decision-making quality, we devise a hierarchical communication graph that enables better collaboration between EVs and facilitates adaptive charging decisions. Finally, extensive experiments demonstrate that e-Charge significantly outperforms compared methods, achieving higher revenue and task completion ratio across a wide range of parameter settings.

Traffic signal control (TSC) provides a cost-effective approach to alleviating urban traffic congestion without requiring modifications to physical road infrastructure. As a classic multi-agent coordination problem, TSC can be optimized with multi-agent reinforcement learning (MARL). However, understanding the cooperative strategies among traffic signals is challenging due to the spatiotemporal dynamics inherent in distributed traffic signal controllers. Previous MARL approaches for TSC often rely on imprecise and redundant cooperation mechanisms, leading to excessive communication costs and inefficient training. In this study, we propose a novel dynamic and controllable cooperation MARL algorithm (DCoo) to optimize traffic signals coordination for alleviating traffic congestion. DCoo introduces a collaboration scheduling module (CSM) to obtain spatiotemporal collaborative information and incorporates a cooperative hyper-decision network (HDN) to optimize decisions by leveraging the processed information. In particular, the CSM first captures crucial temporal information for cooperation by analyzing the relationships between historical and current observations of intersections. CSM then frames the selection of spatial information as a binary classification problem to effectively expand the scope of collaboration among intersections. Additionally, HDN integrates an attention mechanism and permutation invariance to enhance the scalability of DCoo. We conduct comparative experiments on three real-world traffic datasets, and the results demonstrate that DCoo outperforms comparable algorithms in both training and testing phases.