The LEO (Low Earth Orbit) mega constellation network with beam-hopping scheme has become a promising approach to provide global communications. Most of the research on the beam-hopping scheme design for LEO constellation network has focused on beam-hopping scheduling and power allocation for a single LEO satellite. Few of them considered the cooperation between adjacent satellites covering the overlapping areas. As a result, many satellites illuminate beams to the same areas covered by multiple adjacent satellites wasting the limited onboard resource, and the users located in single coverage areas cannot be served. In this paper, we design a multi-agent reinforcement learning (MARL) based cooperative beam-hopping scheme for the LEO mega constellation network, where the LEO satellites cooperatively schedule the beam-hopping and allocate resource. Specifically, we formulate the beam-hopping problem as a multi-agent model, and propose a Centralized Training and Decentralized Execution (CTDE) algorithm. Various simulations are conducted to evaluate our proposed scheme and the results show that the cooperative beam-hopping outperforms the existing schemes in terms of resource utilization and system throughput.

Federated Learning (FL) has emerged as an unparalleled machine learning paradigm where multiple edge clients jointly train a global model without sharing the raw data. However, sharing local models or gradients still compromises clients’ privacy and could be susceptible to delivery failures due to unreliable communication links. To address these issues, this paper considers a multi-server FL where E edge clients wish to jointly train the global model with the help of H servers while guaranteeing data privacy and meanwhile combating$s\leq H$unreliable links per client. We first propose a hybrid coding scheme based on repetition coding and MDS Coding, such that any$T_{s}$colluding servers cannot deduce any client data besides the aggregated model, and any$T_{e}$colluding clients remain unaware of honest clients’ data. Furthermore, we propose a Lagrange coding with mask (LCM) to ensure more stringent privacy protection that additionally demands that colluding servers possess no knowledge about either the local or global models. Furthermore, we establish lower bounds for both the uplink and downlink communication loads and theoretically prove that the hybrid scheme and LCM scheme can achieve the optimal uplink communication loads under the first and second threat models, respectively. For the second threat model with no straggling link, the LCM scheme is optimal. These demonstrate the communication efficiency, robustness, and privacy guarantee of our schemes.

With the widespread use of satellite communications in low earth orbit (LEO), satellite routing has become an important component in ensuring that information can be transmitted efficiently and reliably. However, the time-varying topology and unstable nature of inter-satellite links (ISLs) make it challenging to design LEO satellite routing that guarantees low end-to-end latency. In this paper, we formulate this problem as an optimization problem that aims to minimize the end-to-end latency. We propose a Two-Hop Sate-Aware and Dijkstra-Aided Deep Reinforcement Learning Satellite Routing method (SDDRL-SR) which utilizes Dijkstra's algorithm for supervised training to accelerate the convergence of the algorithm. In the online execution phase, we design a Two-Hop Convergence method that updates the routing strategy by predicting the future two-hop state of the data. Experimental results show that the proposed SDDRL-SR outperforms the Deep Q-Network (DQN), Dijkstra and random routing in the performance of average delay and packet arriving ratio.

With the unprecedented development of smart mobile devices (SMDs), e.g., Internet-of-Things devices and smartphones, various computation-intensive applications are explosively increasing in ultradense networks (UDNs). Mobile-edge computing (MEC) has emerged as a key technology to alleviate the computation workloads of SMDs and decrease service latency for computation-intensive applications. With the benefits of network function virtualization, MEC can be integrated with the cloud radio access network (C-RAN) in UDNs for computation and communication cooperation. However, with stochastic computation task arrivals and time-varying channel states, it is challenging to offload computation tasks online with energy-efficient computation and radio resource management. In this article, we investigate the task offloading and resource allocation problem in MEC-enabled dense C-RAN, aiming at optimizing network energy efficiency. A stochastic mixed-integer nonlinear programming problem is formulated to jointly optimize the task offloading decision, elastic computation resource scheduling, and radio resource allocation. To tackle the problem, the Lyapunov optimization theory is introduced to decompose the original problem into four individual subproblems which are solved by convex decomposition methods and matching game. We theoretically analyze the tradeoff between energy efficiency and service delay. Extensive simulations evaluate the impacts of system parameters on both energy efficiency and service delay. The simulation results also validate the superiority of the proposed task offloading and resource allocation scheme in dense C-RAN.

Abstract In delay and disruption tolerant networks, the contacts among nodes are intermittent. Because of the importance of data access, providing efficient data access is the ultimate aim of analyzing and exploiting disruption tolerant networks. Caching is widely proved to be able to improve data access performance. In this paper, we consider caching schemes for broadcasting and multicasting to improve the performance of data access. First, we propose a caching algorithm for broadcasting, which selects the community central nodes as relays from both network structure perspective and social network perspective. Then, we accommodate the caching algorithm for multicasting by considering the data query pattern. Extensive trace‐driven simulations are conducted to investigate the essential difference between the caching algorithms for broadcasting and multicasting and evaluate the performance of these algorithms. Copyright © 2016 John Wiley & Sons, Ltd.

Asynchronous system provides great performance improvement for wireless ad hoc networks, such as anti-jamming, collision reduction and device simplification. Nevertheless, new media access and routing protocols are required to assist the asynchronous system, e.g., a neighbor discovery algorithm, which is the first step in the initialization of wireless ad hoc networks. In the past few years, a number of algorithms have been proposed for neighbor discovery. However, most of them only consider synchronous system and cannot work efficiently in asynchronous system. In this paper, firstly, we propose an analytical model for an 1-way asynchronous system in wireless ad hoc networks with directional antennas. Then, we compare the time-slot consumption in asynchronous system to complete the neighbor discovery process with that in synchronous system. Finally, in order to improve the performance of the neighbor discovery process, we extend the 1-way asynchronous discovery algorithms to a 2-way asynchronous discovery algorithm. To the best of our knowledge, this is the first practical analytical model of 2-way asynchronous neighbor discovery algorithm with directional antennas.