TDMA scheduling for event-triggered data aggregation in irregular wireless sensor networks

Multi-hop wireless sensor networks often use a TDMA schedule to collect data periodically from multiple locations within a large area. If the measurements from neighboring sensors are cross-correlated, they can be aggregated and compressed as they travel to the data sink. In order for data aggregation to occur quickly, the TDMA schedule must arrange time slot assignments in a certain order. The existing scheduling protocols cannot quickly obtain a schedule with this order, high concurrency, and no collisions. We propose a distributed TDMA scheduling protocol for data aggregation called DATP. In DATP, the sensor nodes transmit dummy packets in order to determine whether they can tolerate the interference from the other nodes that are assigned the same time slot. In this way, time slot allocations are empirically verified to be collision-free. In contrast, the existing protocols obtain schedules with collisions because they use unrealistic interference models such as neglecting interference generated more than two hops away. Furthermore, our simulations reveal that DATP achieves similar concurrency and lower execution time than comparable protocols. These simulations are executed for different network sizes, node densities, and data compression models. In addition, we show that, in networks with fluctuating links, DATP’s main advantage is its execution speed.