Abstract: Due to its benefits of enbancing the spectrum utility in information theory, interference alignment (IA) is quickly adopted in multi-hop wireless networks. The constraints of "node activity in the existing throughput optimization model with IA are insufficient for converging to the optimal solution; the feasibility of IA is not involved. Thus, some measures were taken to compensate for the problems above.The longitudinal constraints of node activity were presented, and the feasibility condition of IA was discussed by the means of setting constraints on the data rates and the macimal number of interfering links at each node. The simulation results demonstrate that the feasible region of the modified optimization mode would be contracted with the longirudinal constraints. Meanwhile, IA feasibility would be guaranteed.

Key words: multi-hop network, interference alignment, mixed integer programming, throughput optimization