In this paper, we investigate user grouping for cooperative scheduling in a two-cell network. When the number of transmitters grows large, the complexity of the Hungarian algorithm optimum for user pairing becomes unaffordable in real-time systems. We consider user grouping algorithms maximizing the network sum rate in cells with a massive number of terminals and/or sensors. We provide a suboptimal user grouping algorithm which substantially reduces complexity compared to the optimum Hungarian algorithm with negligible capacity degradation. Surprisingly, the proposed algorithm outperforms the greedy algorithm with a considerable lower complexity.