Abstract:The low Earth orbit (LEO) satellite edge computing paradigm provides remote sites with flexible, reliable, and scalable edge computing capabilities. Characterized by the orbital motion patterns and harsh space environments, the LEO satellite edge computing faces unique security challenges in terms of the secure collaboration of multiple satellites and the intellectual property protection of models. Under the unique space environment and security demands, we propose a secure satellite edge computing framework in this paper. By taking a remote electricity line outage identification use case as an example, our framework first achieves the secure delegation of the line outage identification task among multiple satellites, which is realized through a secure query $(\mathsf {SQuery})$ scheme to check the availability of the target time slot. Meanwhile, we also design a SHE-enabled secure inner-product encryption ($\mathsf {SSIPE}$) protocol, to achieve the secure multinomial logistic regression (MLR) based line outage identification on-orbit. To reduce the complexity brought by the computationally intensive homomorphic multiplication between two ciphertexts, we further grasp the idea and design a “divide-and-conquer” based secure query ($\mathsf {DSQuery}$) scheme, which converts this homomorphic multiplication operation between ciphertexts into the homomorphic addition operation. As far as we know, this is the first scheme investigating the secure task delegation among different satellites on-orbit. Besides, detailed security analyses are performed to demonstrate the security properties of confidentiality and authentication. In performance evaluations, we test and compare the computational and communication overhead of our scheme and other straightforward schemes. Simulation results show that the $\mathsf {DSQuery}$ scheme greatly reduces the computational cost, which saves the stringent on-orbit computation resources of LEO satellites.

In-Orbit Computation and Security Authentication for Satellite-Ground Twin Networks

Computation and Privacy Protection for Satellite-Ground Digital Twin Networks

Stackelberg Game Based Secure Transmission Strategy for Cognitive Satellite Terrestrial Networks.

Deep Reinforcement Learning-Based Joint Satellite Scheduling and Resource Allocation in Satellite-Terrestrial Integrated Networks

Dynamic Data Offloading for Massive Users in Ultra-dense LEO Satellite Networks Based on Stackelberg Mean Field Game

Distributed Data Offloading in Ultra-Dense LEO Satellite Networks: A Stackelberg Mean-Field Game Approach

Adaptive Multi-Layer Deployment for A Digital Twin Empowered Satellite-Terrestrial Integrated Network

Dynamic Anti-Jamming Strategy in SIoT: A Stackelberg-Matching Game Approach

A Hierarchical Game Framework for Win-Win Resource Trading in Cognitive Satellite Terrestrial Networks

A Game Theory-Based Distributed Downlink Spectrum Sharing Method in Large-Scale Hybrid satellite Constellations

Security-Sensitive Task Offloading in Integrated Satellite-Terrestrial Networks

Toward Multi-Layer Networking for Satellite Network Operations

A Sharded Blockchain-Based Secure Federated Learning Framework for LEO Satellite Networks

A Secure Satellite-Edge Computing Framework for Collaborative Line Outage Identification in Smart Grid

Dual Network Computation Offloading Based on DRL for Satellite-Terrestrial Integrated Networks

Secure and Efficient Token Based Roaming Authentication Scheme for Space-Earth Integration Network

Satellite Based Computing Networks with Federated Learning

On-board Federated Learning for Satellite Clusters with Inter-Satellite Links

Load-Aware Network Resource Orchestration in LEO Satellite Network: A GAT-Based Approach

Orbital collaborative learning in 6G space-air-ground integrated networks

Deep Reinforcement Learning for Privacy-Preserving Task Offloading in Integrated Satellite-Terrestrial Networks