Wireless sensor nodes are usually deployed in remote locations for various applications that require monitoring of certain interesting events. Due to this remote operational feature the longevity of the sensor node's lifetime has been a primary concern. Although the sensor nodes available today may be equipped with rechargeable batteries, the minimal energy capacity of such batteries and low recharge rates degrade the sensor's lifetime and achievable performance. Hence, operational algorithms are needed to guarantee high performance with efficient utilization of energy available. In this thesis, considering temporally correlated event phenomena, the important question answered is: "How long should the sensor sleep, and for how long should the sensor stay active?". To achieve this, a sensor activation/deactivation algorithm has been developed that achieves high performance with efficient energy utilization. A sensor loses energy predominantly because of redundant transmissions of sensed data. To avoid this, a sensor was modeled to transmit only the changes sensed in the event-occurrence process, referred to as Crucial Edges or Transitions. In addition, the system model allows the transmission of transitions that are detected late. Several intuitive decision-making policies were compared and the results compared in order to determine the best policy for this problem. This policy was later analyzed usingMarkov chain analysis techniques to derive upper and lower bounds on the achievable performance. The proposed policy achieves high performance under energy balancing constraints, and is deterministic, simple and easy to implement on a sensor node.