Abstract: In order to solve the problem of uneven distribution of random deployment of traditional sensor networks, cuckoo search algorithm (CS) was proposed to optimize node deployment. In order to improve the global optimization performance of CS algorithm and enhance the deployment optimization ability of sensor nodes, inspired by gradient descent with momentum, root mean square algorithm and Adam optimization algorithm, three improved algorithms, namely Momentum-CS, RMSprop-CS and Adam-CS, were proposed to optimize and adjust the step size control and elimination probability in CS algorithm. Taking network coverage as the optimization target, three algorithms were used to optimize the deployment of water quality monitoring wireless sensor nodes in water area with length and width of 100 meters. Simulation results showed that the Adam-CS algorithm could achieve a higher network coverage of 90.35% in fewer iterations, which was of practical significance for guiding the deployment of wireless sensor nodes in water environment monitoring.

Key words: wireless sensor, water quality monitoring, cuckoo optimization algorithm, Adam optimization