Resilient communication model for satellite networks using clustering technique

• Published in: Reliability engineering & system safety 2021-11, Vol.215, p.107850, Article 107850
• Main Authors: Geng, Sunyue, Liu, Sifeng, Fang, Zhigeng
• Format: Article
• Language: English
• Subjects: Clustering; Clustering technique; Communication; Data integration; Deep reinforcement learning; Dianthus carthusianorum; Dynamic programming; Energy efficiency; Fault tolerance; Network reliability; Quality of service architectures; Reliability analysis; Reliability engineering; Residual energy; Resilience; Resilient communication; Routing (telecommunications); Satellite networks; Satellites
• ISSN: 0951-8320; 1879-0836
• DOI: 10.1016/j.ress.2021.107850

Satellite networks are regarded as significant communication infrastructures to provide high-quality services. With global coverage and flexible deployment, they are able to transfer data swiftly and reliably over long distances. Smooth communication in case of attacks and damage has been paid more attention in recent years. However, little research has addressed resilient schemes for satellite networks so far. To that end, a resilient communication model applicable to satellite networks is put forward. A cluster architecture is constructed with the goal of stable management, data fusion and fault tolerance. In a cluster, there are three types of satellites, clusterhead, core member and ordinary member. The clusterhead is selected depending on the residual energy and the reliability importance of satellites. Moreover, network resilience is comprehensively evaluated in terms of reliability, availability and quality of service (QoS). On that basis, a cluster-based routing protocol with handshake is proposed for maximizing network resilience within basic constraints. Using dynamic programming and deep reinforcement learning method, the routing protocol not only reduces the computation load but also processes data efficiently. Finally, the simulation results demonstrate that the satellite network with the proposed model achieves resilient communication while maintaining high energy efficiency. •A clustering model to provide stable management, data fusion and fault tolerance.•Resilience assessment of satellite networks in terms of reliability, availability and QoS.•A cluster-based routing protocol to maintain normal communication under disruptions.