A sensor network consists of sensor nodes that are constrained in terms of energy consumption. Our work analyzes different ways to both monitor and reduce energy levels in sensor nodes. Monitoring energy levels values is useful for sensor network management.

Energy consumption is distributed among sensor nodes if the nodes store the same data and take turns in sending back data. The first research contribution proposes a cache coherence scheme that enhances upon the previously proposed work of the FIFO cache replacement scheme and assists in localized spatial correlation of sensed data. During the cache updating process the quality of the channel is assessed. We develop an Interference Induced Failed Communication Table (IIFCT) that keeps an account of the failed communication of neighboring nodes. The table entry forms one metric considered for the routing of the data to the Cluster Head.

Sensor Network Management systems require continuous knowledge of network system parameter data to perform data analysis and correlation. The second research contribution proposes an energy metric that is essentially a message that aggregates the residual energies of sensor nodes within a certain bounded region. This Energy metric assists in monitoring network characteristics (residual energy levels). In addition, we also propose a sampling-based methodology to monitor the network system. Sampling-based methodologies have been established to work well with large amounts of data sets. From sampled data, application-based models are created to monitor residual energy levels in an energy efficient way. A time-series model is adopted to predict residual energy levels.

To reduce energy consumption in the application layer, a fourth research contribution proposes information processing algorithms for the sensor network. Simple learning algorithms are proposed to reduce the number of queries sent out by a Cluster Head (sink node). A minimized set of Associated Data and Sensor (ASD) information is kept to assist in the implementation of the learning process.