Performance Analysis of a Non-preemptive Hybrid WSN Protocol in Mobile Environments

摘要

Evolution in Wireless Sensor Networks (WSNs) has allowed new applications that led to an increase in the complexity of communication protocols. Specifically, mobile surveillance applications (where the network is set to monitor mobile objects) require the system to respond in a certain manner so adequate and reliable object tracking and status reporting takes place. Mobile objects can be either independent or controlled by the system, in the former, some sort of transmitter is attached to the mobile entity in order to perform status reporting duties, which, in the case of human beings or specimens may comprise the reporting of vital signs. In view of this, information obtained from mobile objects must be sent with minimal delay and with certain degree of reliability in order to achieve adequate emergency handling when required. Furthermore, mobile environments usually incorporate additional phenomena that creates a multi-event environment, increasing the rate of detection and transmission in the nodes, hence, delaying the report. Hybrid algorithms that allow sensor networks to perform continuous monitoring and event driven applications have proven their ability to enhance performance in different environments where emergency alarms are required while performing a permanent surveillance of the phenomena. Also, these protocols may consider priority assignment in order to reduce report delay and increase reliability in important packets, thus enhancing QoS parameters. In this work, the performance of a non-preemptive hybrid protocol for WSN is studied considering multi-event environments containing mobile targets. Results show that, by using different transmission probabilities, assigned to high and low priority data packets, mobile event reporting delay can be reduced despite the occurrence of other events within the network. As such, a better performance is obtained for critical-time applications where emergency handling is a must.