A wireless sensor network consists of spatially dispersed self-directed sensors to monitor the environment conditions. In probabilistic top-k, a new approach such as expected ranking method is used to get accurate result in finding top-k results and as well as accurate probability. For intercluster processing three algorithm is used namely sufficient-set based, necessary-set based and boundary based algorithm. For reducing transmission cost adaptive algorithm is used and this gives the efficient result in the bounded rounds of communication. This gives least transmission cost and not exceeds two rounds of communication.