For the task assignment of loitering munition swarm cooperatively conducting the reconnaissance，attack and assessment （RAA） tasks，a joint-bidding-based consensus-based bundle algorithm （JCBBA） is proposed by taking into account the constraint of time precedence，the heterogeneous characteristics of target property and the requirement of supporting the multiple munitions to jointly execute an identical task.Munition swarm task assignment of cooperative RAA is first formulated as combinatorial optimization problems.Then based on the structure of CBBA，a task bundle construction mechanism，an individual biding strategy and a marginal benefits computation method considering joint biding are tailored to achieve the fast and robust assignment of task with time precedence constraint in non-center communication networks.Numerical simulations under different scenarios show that JCBBA can achieve reasonable allocation of time precedence tasks while multiple constraints are satisfied，and the comparison tests demonstrate that JCBBA could better trade-off the computation efficiency and optimality with the solving time reduced by approximately 40% compared to consensus-based coalition algorithm.