The brown planthopper (BPH) is a crucial pest of rice in tropical zones like the Mekong Delta of Vietnam. It economically causes severe loss to the rice harvest via direct nutritional depletion. Many studies address the BPH surveillance by using networks of wireless sensors that are mounted on light traps. However, these approaches have not been confirmed as effective deployment due to inoperative light traps’ locations. The problem is that the geographical area of towns is not identical, leading to unnecessary redundancy of sensors and light traps. Our aim in this article is to optimize the locations of BPH sensor networks by utilizing cellular automata and honeycomb architecture which have not been affected by the spatial characteristic geographically. The authors have made several contributions regarding the mentioned problem by (i) quantitatively proving that the deployment cost of BPH sensor networks is significantly reduced, and consequently (ii) optimizing the BPH sensor network. Therefore, the appropriate configuration of the network is maintained in any circumstances. The experiments have been performed on BPH surveillance networks in Hau Giang, a substantial rice province in the Mekong Delta of Vietnam.