Dynamic Resource Allocation for Multi-Satellite Cooperation Networks: A Decentralized Scheme Under Statistical CSI

The densification of low-Earth orbit (LEO) constellations and the development of inter-satellite links enable the cooperative transmission of multiple satellites. This paper investigates a decentralized resource allocation strategy for multi-satellite cooperation networks under statistical channel state information (CSI). Our formulation aims at minimizing the total matching error between non-uniform traffic requests and achievable throughput by jointly optimizing transmit power and beam illumination pattern. The original problem is a stochastic mixed-integer quadratic programming with the long-term accumulation nature, whose stochastic objective is first addressed by incorporating the outage probability constraint. Then we decompose the deterministic problem into a series of single-timeslot subproblems on the greedy basis, each of which is further decomposed for each satellite following the decentralized decision principle with a limited number of inter-satellite interactions required to obtain the overall resource allocation. Finally, the optimization problem at each satellite is solved by sequential quadratic programming (SQP)-based alternating optimization algorithm that iterates between the transmit power allocation and beam illumination design. Numerical results show the effectiveness of the proposed decentralized algorithm in reducing the total matching error by at least 27% over other baseline approaches.