Worst delay perception cross-layer optimization method in wireless sensor network

The invention discloses a worst delay perception cross-layer optimization method in a wireless sensor network and relates to the field of wireless communication transmission technologies. The method comprises the following steps that nodes in the network perform corresponding decision behaviors according to the queue states of a current slot time data queue and a constructed virtual queue, and further update the queue state of the current time slot; source nodes carry out optimized decisions on the data collecting behaviors according to the data queue state and the virtual auxiliary queue state; the source nodes and intermediate nodes carry out optimized decisions on the data packet dropout behaviors and the route mobilization and data transmission behaviors according to the current time slot data queue state and the virtual delay queue state. By means of multiple times of iteration, the queue state of the wireless sensor network is gradually stabilized. Meanwhile, the method can make the handling capacity of the wireless sensor network effectiveness tend to be optimal, make the integral time average data packet dropout quantity of the network tend to be zero, and provide the worst transmission delay upper bound for data of each session stream in each node of the network.