Load Balanced Handover Planning for LEO Satellite Communication Networks

To ensure uninterrupted service for a user, frequent handovers are essential in a LEO satellite network. Since satellites follow predictable orbits, the movement pattern of a satellite is deterministic. Exploiting the predictability of satellite locations, we devise a handover scheduling plan that incorporates load balancing considerations in satellite networks. To reduce computational complexity, we organize users in close proximity into groups and formulate handover planning based on these user groups. We approach the problem by treating it as a routing problem in an auxiliary graph, which represents the candidate satellites capable of providing qualified connectivity for a user group. We formulate the handover planning problem as two concatenated Integer Linear Programming (ILP) problems. The objective function of the first problem aims to minimize the traffic on the most heavily loaded satellite. The load on the most congested satellite, obtained from the first problem, serves as a constraint in the second problem. This constraint is employed to set an upper limit on the load for each satellite. The objective function of the second problem is to minimize the total number of handovers. The numerical results demonstrate that our approach significantly reduces the load on the most congested satellite compared to methods that use signal strength and longest service time as handover criteria. Our handover planning results in only a slight increase in the total number of handovers compared to the longest service time strategy.