Title :
Improving packet reception performance in high traffic sensor networks
Author_Institution :
Comput. Sci. Dept., State Univ. of New York, Stony Brook, NY
Abstract :
The increasing capabilities of software and hardware in sensor networks enable a broad range of data-intensive applications such as sensor databases, image and video processing, real-time object localization and tracking, and pervasive health monitoring. These applications require high data rates to complete node collaboration tasks. Thus, a reliable and efficient data receiving mechanism is needed for these emerging applications. Existing operating systems dedicated to sensor networks, however, show little support for high traffic data rate; reliability and congestion control are pushed to upper networking layers. These strategies increase the overhead of communication between sensor nodes because of state transitions through message exchanges. Due to energy constraints, the overhead imposed by these protocols will also reduce the lifetime of the network. This paper proposes a simple, but efficient data receiving strategy at the kernel layer which can be applied to most sensor OSes, including TinyOS and Contiki. We queue a set number of packets into a receive buffer (RBUFF) and return the message pointer immediately to the radio driver underneath the network layer. Thus, incoming packets are not dropped if the upper layer protocol can not handle them quickly enough; outstanding packets are queued in the RBUFF.
Keywords :
packet radio networks; queueing theory; telecommunication congestion control; telecommunication traffic; congestion control; data receiving mechanism; high traffic sensor network; packet reception; Application software; Communication system traffic control; Hardware; Image databases; Image sensors; Monitoring; Protocols; Telecommunication network reliability; Telecommunication traffic; Traffic control;
Conference_Titel :
Pervasive Computing and Communications, 2009. PerCom 2009. IEEE International Conference on
Conference_Location :
Galveston, TX
Print_ISBN :
978-1-4244-3304-9
Electronic_ISBN :
978-1-4244-3304-9
DOI :
10.1109/PERCOM.2009.4912805