Wireless sensor networks are currently deployed in many areas, particularly for surveillance related applications. Sensors have very limited energy and processing capabilities, hence, it becomes necessary to introduce energy efficient algorithms to maximize the lifetime of a sensor node. We propose a new scheduling scheme based on Discrete Time Markov chain models used in genetics for DNA evolution prediction. The proposed scheduler uses a single control parameter to control state changes in order to obtain a compromise between network lifetime and throughput. We discuss the design of such a Discrete Time Markov chain based scheme and compare it to a standard approach in terms of node throughput and lifetime of entire network. Finally, we show the effectiveness of this scheme by simulating various network topologies in a realistic sensor network. Our observations show that just after 75% of simulation steps 90% more nodes are alive with the proposed scheduler. The residual battery power is 82% more and the packet reception rate is increased by 51% for the entire network when compared to the standard approach.