The Impact of Storage Capacity Usage and Predictable Contact Schedule on Dynamic Routing for Opportunistic Deep Space Information Networks

In this paper, the impact of storage capacity usage and predictable contact schedule on dynamic routing is studied for the Opportunistic Deep Space Backbone Layer (OppDSBL) of the Deep Space Information Networks. The hypergraph model is firstly constructed to represent the OppDSBL for the sake of the improvement towards the network connectivity. Also, a polynomial time algorithm of dynamic routing in the OppDSBL is proposed, which realizes joint routing selection, storage capacity usage and predictable contact schedule simultaneously. The proposed algorithm depends on two novel models, i.e., the dynamic storage capacity usage model (DSCUM) and the predictable contact schedule model (PCSM). In the DSCUM, a general infinite-horizon differential game model for dynamic storage capacity usage of joint node is implemented based on two cost functions, which indicates the reduction as well as the consumption of storage capacity. Moreover, an optimization problem is further designed by taking into account the impact of selfish behavior on the dynamic storage capacity usage, and a set of optimal strategies to the noncooperative dynamic storage capacity usage is also obtained. In the PCSM, the predictable contact interruption probability is explicitly quantified to estimate the likelihood of link outage. In addition, the predictable contact schedule metric is further formulated to describe the contact relationship among different isolated islands in hypergraph model. Theoretical analysis and numerical results demonstrate the effectiveness and feasibility of the proposed dynamic routing algorithm.