From Rejection to Re-Accommodation: A Generate-and-Negotiate Framework for Conflict-Free U-space

High-density very low level operations frequently lead to flight plan rejections under strategic deconfliction ser- vices that preserve trajectories and adjust the take-off time window. This study addresses the problem through alternative generate based on a Single-Interval alternative solving model: for each previously rejected flight plan, a continuous conflict- free alternative requested launch window is computed within a bounded range, allowing drone operators to select feasible departures without trajectory changes while maintain the safety requirements. The method relies on constraint-based scheduling with separation, capacity, and constraints, and optimizes either acceptance or temporal flexibility subject to operational lim- its. Results on representative urban-corridor scenarios indicate substantial reductions in rejections and large gains in available flexibility at low computational cost. The outputs are packaged into a global solution table that aggregates globally feasible com- binations of the alternative requested launch window to ensure that subsequent selections remain conflict-free. A multi-agent negotiation framework can consume the global solution table to clear among solutions using sealed-bid. Overall, Single-Interval alternative solving model coupled with deep reinforcement learn- ing based multi-agent negotiation framework converts rejections into explainable, auditable re-accommodations while remaining compatible with current strategic deconfliction workflows.