Efficient Data Gathering in Intermittent Connected Mobile Sensor Networks

Intermittent Connected Mobile Sensor Networks (ICMSN) have emerged in recent studies, e.g. underwater sensor networks, wildlife tracking and human-oriented flu virus monitoring. In these networks, conventional routing protocols in WSN are not suitable for data gathering application in ICMSN since packets will be dropped when no complete route to the sink is available. In this paper, we propose an efficient data gathering scheme for ICMSN, which is composed of forwarding protocol and buffer management. Message forwarding makes decision on which sensors are qualified for next hop based on nodal delivery probability, which signifies the likelihood that the sensor can deliver the message to the sink successfully based on the nodal mobility pattern and residual energy. Buffer management consists of two related components: queuing mechanism and purging mechanism. The former determines priority of messages to transmit based on the importance factor, which indicates importance of the messages and QoS requirement of transmission. The latter utilizes a death vector generated by the sink to purge useless messages that have been delivered from the network. Our experimental results show that the proposed data gathering scheme supports the different transmission requirement of messages and achieves the good performance tradeoff between delivery ratio, delay and resource consumption.