Data Aggregation Based Reliable Transport Method In Wireless Sensor And Actor Networks

Wireless sensor and actor network is a new self-organizing wireless network. It is composed of a large number of sensor nodes and several mobile actors. The mobile actors move in a large area to execute missions, such as data collection. The sensor nodes monitor the area and then forward their readings to the actors. In this paper, we study the data reliable transport problem of a special high-speed mobile sensor and actor network which uses the high speed train to collect the monitoring data of tunnels and bridges.In recent years, research on using wireless sensor networks (WSNs) for structural health monitoring (SHM) has attracted increasing attention. In such applications, the sensor nodes are deployed in the tunnel or the bridge to detect the possible structure damage. Then the detected data are reported to the base station which is always in the city far away from the tunnel or the bridge. In this paper, we thus propose to leverage high-speed rail as mobile sink to assist data collection and delivery.Due to high density in the network topology, sensor observations have spatial and temporal correlation. We focus on discussing how to use this correlation to guarantee the reliability of the high-speed actor (train) data collection. This paper provided a two-phase data collection scheme which contained sensor aggregation phase and actor aggregation phase. In the sensor aggregation phase, the WSANs transmit the data to a sink node and then aggregate the data using the spatial correlation. In the actor aggregation phase, when the high-speed train passing by the tunnel or the bridge, the sink node transmits the data to the train and then the information processing center on the train aggregates the data with consideration of the temporal correlation. Also because of the high speed mobility of the train, the communication link between the sink and the train may be fail, so we need to estimate the lost packet.Then we study the transport capacity of data-gathering WSNs with correlation aware aggregation in the sensor aggregation phase. A grid partition method is proposed to divide the network into sectors of the same size, and then correlation aware aggregation is presented to analyze the throughput of the network. Finally we propose a data distortion analysis model of the whole sensor and actor networks, and then estimate spatiotemporal distortion to denote the data collection's reliability/fidelity. Through extensive simulation, we discuss several key elements that will affect the reliability of the high-speed sensor and actor networks data collection, and then give the improved scheme.