In order to improve lifetime of sensors, it is important to obtain a smaller partial set of working sensor nodes for full coverage in wireless sensor networks. This paper presents a decentralized self-configured method for determining active nodes required to ensure complete coverage. A two-way coverage mechanism identifies blind/overlapping areas and activates/deactivates sensor nodes as required to maintain the coverage. The mechanism was implemented successfully using the network simulator NS-2. The performance was compared to that provided by three other alternative approaches, including OTTAWA, PEAS, and OGDC. The results show that the two-way algorithm is network scalable and coverage efficient.