Directed diffusion for wireless sensor networking Export

@article{Intanagonwiwat2003Directed, abstract = {Advances in processor, memory and radio technology will enable small and cheap nodes capable of sensing, communication and computation. Networks of such nodes can coordinate to perform distributed sensing of environmental phenomena. In this paper, we explore the directed diffusion paradigm for such coordination. Directed diffusion is datacentric in that all communication is for named data. All nodes in a directed diffusion-based network are application-aware. This enables diffusion to achieve energy savings by selecting empirically good paths and by caching and processing data in-network. We explore and evaluate the use of directed diffusion for a simple remote-surveillance sensor network.}, address = {Piscataway, NJ, USA}, author = {Intanagonwiwat, Chalermek and Govindan, Ramesh and Estrin, Deborah and Heidemann, John and Silva, Fabio}, citeulike-article-id = {315814}, citeulike-linkout-0 = {http://portal.acm.org/citation.cfm?id=638335}, citeulike-linkout-1 = {http://dx.doi.org/10.1109/TNET.2002.808417}, citeulike-linkout-2 = {http://dx.doi.org/http://dx.doi.org/10.1109/TNET.2002.808417}, citeulike-linkout-3 = {http://dx.doi.org/10.1109/TNET.2002.808417}, comment = {in WSNs - data centric control/comms model (not point-to-point like IP) It is not unreasonable to expect the following features in a future sensor node: A matchbox sized form factor, battery power source, an power-conserving processor clocked at several hundred Mhz, program and data memory amounting to several tens of MBytes, a radio modem that employs some form of diversity coding [10], and an energy efficient MAC layer based on, for example, TDMA Compared to engineered sensor systems which 1) complex targets, high processing 2) or simple sensed time series Could use IP/adhoc routing, but data centric model we can obtain better performance for the specific task. Still a fairly specific task. hmmmm. Data is named using attribute pairs. Assume positional information. Specify a task as an "interest". Every node between a source and sink contains an interest gradient field for every neighbour. Every node appears as a source for their neighbours, passing data on in a manner described by their interests. Initial flooding of network to establish gradients. Then reinforce gradients according to returned data. Gives local filtering and data combination. Quite elegant - local, distributed communication, redundant data initially, Simulation using ns, compare to 1) flooding - worst case energy consumption 2) omniscient multicast (shortest path multicast scheme between source and sink) - IP equivalent Better energy consumption than omniscient multicast- due to in network aggregation. Using IEEE802.11 MAC - not optimal, rather use TDMA MAC.}, doi = {10.1109/TNET.2002.808417}, issn = {1063-6692}, journal = {IEEE/ACM Trans. Netw.}, keywords = {data-centric-routing, directed-diffusion, protocol-design, simulation, wsn}, month = {February}, number = {1}, pages = {2--16}, posted-at = {2008-11-24 15:58:07}, priority = {0}, publisher = {IEEE Press}, title = {Directed diffusion for wireless sensor networking}, url = {http://dx.doi.org/10.1109/TNET.2002.808417}, volume = {11}, year = {2003} }

TY - JOUR ID - Intanagonwiwat2003Directed L3 - citeulike-article-id:315814 N2 - Advances in processor, memory and radio technology will enable small and cheap nodes capable of sensing, communication and computation. Networks of such nodes can coordinate to perform distributed sensing of environmental phenomena. In this paper, we explore the directed diffusion paradigm for such coordination. Directed diffusion is datacentric in that all communication is for named data. All nodes in a directed diffusion-based network are application-aware. This enables diffusion to achieve energy savings by selecting empirically good paths and by caching and processing data in-network. We explore and evaluate the use of directed diffusion for a simple remote-surveillance sensor network. IS - 1 JF - IEEE/ACM Trans. Netw. SN - 1063-6692 AD - Piscataway, NJ, USA EP - 16 TI - Directed diffusion for wireless sensor networking PB - IEEE Press VL - 11 SP - 2 KW - data-centric-routing KW - directed-diffusion KW - protocol-design KW - simulation KW - wsn N1 - in WSNs - data centric control/comms model (not point-to-point like IP) It is not unreasonable to expect the following features in a future sensor node: A matchbox sized form factor, battery power source, an power-conserving processor clocked at several hundred Mhz, program and data memory amounting to several tens of MBytes, a radio modem that employs some form of diversity coding [10], and an energy efficient MAC layer based on, for example, TDMA Compared to engineered sensor systems which 1) complex targets, high processing 2) or simple sensed time series Could use IP/adhoc routing, but data centric model we can obtain better performance for the specific task. Still a fairly specific task. hmmmm. Data is named using attribute pairs. Assume positional information. Specify a task as an "interest". Every node between a source and sink contains an interest gradient field for every neighbour. Every node appears as a source for their neighbours, passing data on in a manner described by their interests. Initial flooding of network to establish gradients. Then reinforce gradients according to returned data. Gives local filtering and data combination. Quite elegant - local, distributed communication, redundant data initially, Simulation using ns, compare to 1) flooding - worst case energy consumption 2) omniscient multicast (shortest path multicast scheme between source and sink) - IP equivalent Better energy consumption than omniscient multicast- due to in network aggregation. Using IEEE802.11 MAC - not optimal, rather use TDMA MAC. AU - Intanagonwiwat, Chalermek AU - Govindan, Ramesh AU - Estrin, Deborah AU - Heidemann, John AU - Silva, Fabio PY - 2003/February// UR - http://dx.doi.org/10.1109/TNET.2002.808417 ER -