Abstract
The integrated satellite-terrestrial networks (STNs) aim to provide global connectivity and support heterogeneous services, including enhanced mobile broadband (eMBB) and massive machine-type communication (mMTC). Each service request requires a series of virtual network functions (VNFs) to be deployed consecutively. The VNFs are mapped on nodes that constitute a path where a request is mapped. Provisioning multiple slices through satellite networks is challenging due to limited storage and computation resources. In addition, there are dynamic changes in the satellite positions that cause frequent variations in the topology. For requests lasting more than one time frame, a handover can be performed at the beginning of the next time frame. Handover implies overall reconfiguration, which induces significant computation costs in satellite networks. Therefore, in this article, we propose a path and VNF mapping strategy with selective handover while considering dynamic changes in the satellite topology and the limitation of available resources. We formulate a mathematical model based on Binary Integer Linear Programming (BILP), aiming to maximize the served requests. To reduce the time complexity of the model, we solve it using an iterative algorithm based on successive convex approximation (SCA). We refine the solution of the SCA after binary recovery using VNF and path mapping algorithm. For the mapped multi-frame requests in the current time frame, the resources are reserved in the nodes and links for the next time frame if the current routing path is available for the remaining duration of the request. The simulation results certify the performance of the proposed technique with a significant improvement in the served request percentage compared to previous works in the literature, while also reducing the number of handovers.