Low earth orbit (LEO) satellite networks, which integrate with communication, sensing, and computing capabilities, have emerged as a promising approach to improve the performance of service quality and network utility. In that case, cooperative edge computing among several satellites and heterogeneous onboard resource management is viewed as a significant challenge, particularly since the task of data processing involves multiple functions with complex serial and parallel dependencies that need to be scheduled and executed in a particular order. Therefore, in this paper, the heterogeneous resource management for complex task offloading in LEO satellite networks is investigated by introducing directed acyclic graph (DAG). The problem of DAG-based task offloading is formulated to minimize the joint deployment costs in terms of computing, network bandwidth, and service delay. A neighborhood-based breadth first search (N-BFS) approach is proposed to obtain sub-optimal solutions of VNF placement and data routing in an acceptable computation complexity. The experiments with three baselines of Viterbi, BFS, and Gurobi are conducted to evaluate the performance of N-BFS. Simulation results demonstrate that the N-BFS approach is effective and efficient for solving the DAG-based task offloading problem by trade-off between quality of solution and time complexity.

With the rapid development of large low earth orbit (LEO) satellite constellations, satellite edge computing is an emerging topic to provide computing services for Internet of Things (IoT) users, which are not in the coverage of terrestrial networks. For computation offloading in satellite edge computing, it is still challenging to allocate the network resources on-demand for IoT users to improve service experience while reducing energy consumption, since user tasks may be offloaded between different satellites by inter-satellite links (ISLs). In this paper, we study the joint optimization problem of computation offloading and resource allocation in cooperative satellite edge computing. Then, a hierarchical dynamic resource allocation (HDRA) algorithm for computation offloading is proposed by introducing breadth first search (BFS) and greedy to tackle the problem, the aim is to minimize service delay and energy consumption jointly. We conduct the experiments to evaluate the performance of the proposed HDRA algorithm, compared with two baselines of BFS-PSO and Gurobi. Experimental results show that the proposed HDRA algorithm can address the formulated problem effectively in satellite edge computing and obtain the results of computation offloading and resource allocation in a low running time.

The deployment of satellite mobile edge computing (SMEC) incorporating artificial intelligence (AI) in low Earth orbit (LEO) constitutes satellite edge intelligence (SEI), which is promising to achieve autonomous processing of space missions on board driven by massive data. However, individual satellites with constrained resources and insufficient samples learn inefficiently, while the spatio-temporal constraints of large-scale LEO networks make collaborative training difficult. In this paper, a leader federated learning (FL) architecture for distributed SEI (SELFL) is proposed. By evaluating the connectivity and load of the dynamic constellation, the global and local parameters of the shared AI model are transmitted and updated continuously between the elected leader and other follower satellites based on the established inter-satellite link, which realizes efficient self-evolution of SELFL independent of the ground. Also we introduce a deep reinforcement learning-based resource allocation strategy for SELFL, which leverages the distributed proximal policy optimization (DPPO) to optimize the computing capability and transmit power of satellites for accelerating FL and reducing energy consumption. This method not only updates stably utilizing adaptive learning steps, but also improves sample efficiency with multiple parallel workers. The simulation results demonstrate the proposed SELFL optimization scheme effectively reduces the total energy consumption and training time by ensuring the AI model accuracy, and outperforms the benchmark algorithms.

Enabling a satellite network with edge computing capabilities can complement the advantages further of a single terrestrial network and provide users with a full range of computing service. Satellite edge computing is a potentially indispensable technology for future satellite-terrestrial integrated networks. In this article, a three-tier edge computing architecture consisting of the terminal–satellite–cloud is proposed, where tasks can be processed at three planes and intersatellites can cooperate to achieve on-board load balancing. Facing varying and random task queues with different service requirements, we formulate the objective problem of minimizing the system energy consumption under the delay and resource constraints, and jointly optimize the offloading decision, communication, and computing resource allocation variables. Moreover, the distribution of resources is based on the reservation mechanism to ensure the stability of the satellite-terrestrial link and the reliability of computation process. To adapt to the dynamic environment, we propose an intelligent computation offloading scheme based on the deep deterministic policy gradient (DDPG) algorithm, which consists of several different deep neural networks (DNNs) to output both discrete and continuous variables. Additionally, by setting the selection process of legal actions, the simultaneous decisions on offloading locations and allocating resources under multitask concurrency is realized. The simulation results show that the proposed scheme can effectively reduce the total energy consumption of the system by ensuring that the task is completed on demand, and outperform the benchmark algorithms.

Satellite networks, as a supplement to terrestrial networks, can provide effective computing services for Internet of Things (IoT) users in remote areas. Due to the resource limitation of satellites, such as in computing, storage, and energy, a computation task from an IoT user can be divided into several parts and cooperatively accomplished by multiple satellites to improve the overall operational efficiency of satellite networks. Network function virtualization (NFV) is viewed as a new paradigm in allocating network resources on-demand. Satellite edge computing combined with the NFV technology is becoming an emerging topic. In this paper, we propose a potential game approach for virtual network function (VNF) placement in satellite edge computing. The VNF placement problem aims to minimize the deployment cost for each user request, furthermore, we consider that a satellite network should provide computing services for as many user requests as possible. We formulate the VNF placement problem as a potential game to maximize the overall network payoff and analyze the problem by a game-theoretical approach. We implement a decentralized resource allocation algorithm based on a potential game (PGRA) to tackle the VNF placement problem by finding a Nash equilibrium. Finally, we conduct the experiments to evaluate the performance of the proposed PGRA algorithm. The simulation results show that the proposed PGRA algorithm can effectively address the VNF placement problem in satellite edge computing.

In this paper, we investigate the service chaining placement problem for user requests in satellite ground station networks with minimum energy cost, which consists of server energy, switch energy, and link energy. We build the Server-Switch-Link energy model and formulate the energy optimization problem as an integer nonlinear programming problem. To address this problem, we implement a prediction-aided Greedy (PA-Greedy) algorithm depending on satellite mission planning in satellite control centers. We conduct the experiments to evaluate the proposed energy model and PA-Greedy algorithm in Fat-Tree networks, and compare the performance with the baseline Greedy algorithm and two energy models of Server-Link and Server. In a Fat-Tree network with 16 servers, the proposed Server-Switch-Link energy model with the PA-Greedy algorithm can reduce energy consumption by 17.28% when compared with the baseline Greedy algorithm, and outperform these Server-Link and Server energy models by 47.10% and 62.91%, respectively.

As the increase in satellite number and variety, satellite ground stations should be required to offer user services in a flexible and efficient manner. Network function virtualization (NFV) can provide a new paradigm to allocate network resources on-demand for user services over the underlying network. However, most of the existing work focuses on the virtual network function (VNF) placement and routing traffic problem for enterprise data center networks, the issue needs to further study in satellite communication scenarios. In this article, we investigate the VNF placement and routing traffic problem in satellite ground station networks. We formulate the problem of resource allocation as an integer nonlinear programming (INLP) model and the objective is to minimize the link resource utilization and the number of servers used. Considering the information about satellite orbit fixation and mission planning, we propose location-aware resource allocation (LARA) algorithms based on Greedy and IBM CPLEX 12.10, respectively. The proposed LARA algorithm can assist in deploying VNFs and routing traffic flows by predicting the running conditions of user services. We evaluate the performance of our proposed LARA algorithm in three networks of Fat-Tree, BCube, and VL2. Simulation results show that our proposed LARA algorithm performs better than that without prediction, and can effectively decrease the average resource utilization of satellite ground station networks.

In cloud computing, an important concern is to allocate the available resources of service nodes to the requested tasks on demand and to make the objective function optimum, i.e., maximizing resource utilization, payoffs, and available bandwidth. This article proposes a hierarchical multi-agent optimization (HMAO) algorithm in order to maximize the resource utilization and make the bandwidth cost minimum for cloud computing. The proposed HMAO algorithm is a combination of the genetic algorithm (GA) and the multi-agent optimization (MAO) algorithm. With maximizing the resource utilization, an improved GA is implemented to find a set of service nodes that are used to deploy the requested tasks. A decentralized-based MAO algorithm is presented to minimize the bandwidth cost. We study the effect of key parameters of the HMAO algorithm by the Taguchi method and evaluate the performance results. The results demonstrate that the HMAO algorithm is more effective than two baseline algorithms of genetic algorithm (GA) and fast elitist non-dominated sorting genetic algorithm (NSGA-II) in solving the large-scale optimization problem of resource allocation. Furthermore, we provide the performance comparison of the HMAO algorithm with two heuristic Greedy and Viterbi algorithms in on-line resource allocation.

Spinal codes are a new kind of rateless codes and can achieve the shannon capacity over binary symmetric channel (BSC) and additive white gaussian noise (AWGN) channel. This paper describes a self-adaptive wireless communication system for spinal codes. The key points to our design include the integrated methods of coding and mapping for spinal codes, transmission protocols and feedback frame. Finite feedback transmission protocol for the cumulative probability distribution function (CDF) decoding and multi-frame aggregation protocol are introduced to improve the rate of transmitted data for the system. We also propose a mechanism of interference cancellation (IC) to handle the problem of packet loss. Finally, we design and implement this system by using IEEE 802.11. The results of our simulation show that our system is able to operate adaptively to copying with time-varying channel and provide a good performance.