Scheduling and data aggregation for periodic data gathering in wireless sensor networks

Wireless Sensor Networks consist of multiple energy-constrained sensing devices called sensor nodes. Typically, the sensor nodes generate data that must reach a single data sink across multiple hops. The data from neighboring nodes are usually correlated and can be aggregated and compressed inside of the network. This process is referred to as data aggregation and requires the development of data aggregation functions. Before data aggregation functions can be developed, uncompressed measurements need to be collected. To gather such measurements, we propose a TDMA protocol that assigns time slots very efficiently in networks in which the traffic pattern and the topology change slowly. Data aggregation requires significant coordination between the sensor nodes because the packets to be aggregated must lie at the same node at the same time, thereby affecting the routing and MAC layers. At the routing layer, we propose a protocol to quickly obtain a routing tree for data aggregation in a network where sensor nodes sleep for long periods of time. At the MAC layer, we provide a TDMA protocol in which transmissions are arranged in order suitable for data aggregation. Our protocol outperforms the existing protocols in terms of energy consumption, reliability, and spatial reuse. For networks with unreliable links, we propose a protocol to decide which packets to transmit and which packets to discard in order to balance the energy consumption of the nodes while satisfying data fidelity constraints.