A hybrid method for the QoS analysis and parameter optimization in time-critical random access wireless sensor networks

Evolution in electronics has led to the development of complex applications in wireless sensor networks (WSNs), where efficient and swift event reporting is needed. In time-critical applications, achieving an adequate report latency is particularly relevant as it allows a proper reaction from the network to the occurring phenomena. It is evident that mean report latency is insufficient as a QoS indicator for time-critical applications. Instead, high percentiles or the whole distribution are much better suited. In certain applications such as target tracking and positioning, the transmission of a certain number of event packets is required to accurately characterize the occurring phenomena. Building on this, we present a hybrid method for obtaining the probability distribution of report latency in random access (RA) WSN protocols. In this method, the distribution of the number of detecting nodes is obtained by simulation, which then allows us to obtain the desired QoS parameters analytically. In this study, we use our method to obtain and optimize the event report latency and energy consumption in RA WSNs. Results show that modifying the transmission parameters during backoff increases the robustness of RA event reporting and also enhances the performance of the WSN in environments where multiple types of events can be detected.