Protocols for Sensor Networks
   

 
Project
Overview

 
 
BWN-Lab
Personnel

 
 
Project
Description

 
 Testbed

 
 
Publications

 
 
BWN-Lab
Research Projects
You are visitor

since 06/30/2006

Project Description
Sensor Networks' Communication Protocols

Current architectures for Internet and ad hoc wireless networks may not be used for wireless sensor networks.  Here are few of the reasons: 

  • Number of sensor nodes in a sensor network is much higher than ad hoc network. 
  • Sensor nodes experience failures much more frequently than nodes in ad hoc network.
  • Sensor nodes are simpler than nodes in the Internet and ad hoc networks. 
  • Sensor nodes are very limited in power. 
  • The header of a Internet packet is too long for sensor networks, e.g., each node must have a permanent address.
  • The use of acknowledgment packet should be used sparingly.
As a result, a new architecture will:
  • Combine power and routing awareness,
  • Integrate data with networking protocols,
  • Communicate power efficiently through the wireless medium, and
  • Share tasks among neighbors.
Our new protocol stack will address the new architecture's requirements.

Related Work

  • Akyildiz, I. F., Su, W., Sankarasubramaniam, Y., and Cayirci, E.,"A Survey on Sensor Networks," IEEE Communications Magazine, Vol. 40, No. 8, pp. 102-114, August 2002; receives the IEEE Communications Society 2003 Best Tutorial Paper Award, April 2003.
Back to Top


Spatio-Temporal Correlation in Wireless Sensor Networks

Wireless Sensor Networks (WSN) are characterized by the dense deployment of sensor nodes that continuously observe physical phenomenon. Due to high density in the network topology, and the nature of the physical phenomenon, sensor observations are highly correlated in the space and time domains, which constitutes the spatial and temporal correlation. 
  • Spatial Correlation: Typical WSN applications require spatially dense sensor deployment in order to achieve satisfactory coverage. Due to high density in the network topology, spatially proximal sensor observations are highly correlated with the degree of correlation increasing with decreasing internode separation. 
  • Temporal Correlation: Some of the WSN applications such as event tracking may require sensor nodes to periodically perform observation and transmission of the sensed event features. The nature of the energy-radiating physical phenomenon constitutes the temporal correlation between each consecutive observation of a sensor node. The degree of correlation between consecutive sensor measurements may vary according to the temporal variation characteristics of the phenomenon.

These spatial and temporal correlations along with the collaborative nature of the WSN bring significant potential advantages for the development of efficient communication protocols well-suited for the WSN paradigm. In this project, several key elements are investigated to capture and exploit the correlation in the WSN for the realization of advanced efficient communication protocols. A theoretical framework is developed to model the spatial and temporal correlations in sensor networks. Based on this framework, possible approaches are discussed to exploit spatial and temporal correlation for efficient medium access and reliable event transport in WSN, respectively.

Related Work

  • Vuran M. C., Akan O. B., Akyildiz I. F., "Spatio-Temporal Correlation: Theory and Applications Wireless Sensor Networks," Computer Networks Journal (Elsevier Science), vol. 45, no. 3, pp. 245 -259, June 2004.
  • Akyildiz I. F., Vuran M. C., Akan O. B., "On Exploiting Spatial and Temporal Correlation in Wireless Sensor Networks," in Proc. WiOpt'04: Modeling and Optimization in Mobile, Ad Hoc and Wireless Networks, pp. 71 -80, March 2004.
Back to Top


Event-to-Sink Reliable Transport (ESRT)

Reliable detection of event features necessitates an event-to-sink reliability notion in contrast to existing end-to-end reliability (TCP, SRM, PSFQ). This is due to the fact that correlated data flows from several source nodes to a single sink are loss tolerant to the extent that event features are reliably detected. The use of strict end-to-end reliable transport in such a case can lead to over-utilization of scarce sensor resources.
ESRT seeks to achieve reliable event detection with minimum energy expenditure and congestion control. It has been tailored for use in sensor networks with adaptability to dynamic topology, collective identification, energy conservation and biased implementation at the sink.

Related Work

  • Akan, O.B., and Akyildiz, I.F., ``Event-to-Sink Reliable Transport in Wireless Sensor Networks,'' to appear in IEEE/ACM Transactions on Networking, October 2005.
  • Sankarasubramaniam, Y., Akan, O.B., and Akyildiz, I.F., "ESRT: Event-to-Sink Reliable Transport in Wireless Sensor Networks," in Proc. ACM MobiHoc'03, Annapolis, Maryland, USA, June 2003.
Back to Top

Routing Protocol

There may be several sinks in a sensor network. The sinks are gateways between the sensor network and the backbone network, e.g. Internet.  Note that the sink may be in a ground-based site possibly set up by a rapid response team, in an unmanned airborne vehicle or plane, or a low earth orbit satellite.  Depending on the DoD mission, the sinks and sensor nodes may be mobile.  The objective of the routing protocol is to deliver sensed information from the sensor nodes, i.e., the sources, to the appropriate sinks.  Sensed information will be represented by descriptors, which will be fused, i.e. if local neighbors have same descriptors, the descriptors will be combined, before they are routed to the sinks.

The routing protocol must meet the following design targets:

  • They must be power efficient.  Sensor nodes have low power capacity thus, power is a very important issue.  The lifetime of a sensor node ends with the battery.  As a result, redundant transmissions must be as low as possible.
  • It must be reliable.  Sensor nodes will deal with critical data in unreliable wireless environment.
  • Delays must be low.  The sensor network may also be used for real-time sensing.  Thus, delay is an important issue.
  • Power emanation must be low.  In many DoD missions the sensor network must be undetectable.  Thus, the power emanation must be kept low.
Open Research Issues
  • Routing protocols under mobile scenarios is relatively understudied. The basic approach of routing data back along gradients created by interest dissemination fails under dynamic topologies.
  • Note that dynamic topologies can be the result of both node mobility and node failure/temporary power-down. Randomized routing protocols have been shown to adequately deal with the latter problem. However, the problem of neighbor discovery remains for the mobile case. Mobile routing protocols are of interest in several military applications.
Related Work
    Su, W. and Akyildiz, I. F.,"A Stream Enabled Routing (SER) Protocol for Sensor Networks," Med-hoc-Net 2002, Sardegna, Italy, September 2002.   
Back to Top


Topology Control for Geographical Routing

Geographical routing algorithms are considered in sensor networks because of their scalability. In a geographical routing algorithm, the packets are forwarded by a node to its neighbor based on their respective positions. Thus, there is no need to store and maintain routing tables in constrained sensor nodes. Geographical routing algorithms are known to be scalable but their energy efficiency has never been extensively and comparatively studied. 

In this project, a novel analytical framework has been introduced, which allows to analyze the relationship between the energy efficiency of the routing tasks and the extension of the range of the topology knowledge for each node. A wider topology knowledge can improve the energy efficiency of the routing tasks but can also increase the cost of topology information due to signaling packets that each node must transmit and receive to acquire this information, especially in networks with high mobility. It is demonstrated that a limited knowledge of the topology is sufficient to take energy efficient forwarding decisions.

Related Work

  • Melodia, T., Pompili, D., and Akyildiz, I.F., "Optimal Local Topology Knowledge for Energy Efficient Geographical Routing in Sensor Networks,'' in Proc. of IEEE Infocom 2004, Hong Kong, P.R. China, March 2004.
  • Melodia, T., Pompili, D., and Akyildiz, I.F., "On the Interdependence of Distributed Topology Control and Geographical Routing in Ad Hoc and Sensor Networks,'' to appear on JSAC Special Issue on Wireless Ad Hoc Networks, March 2005.
Back to Top

Data Link Protocol: MAC

Wireless Sensor Networks (WSN) are characterized by dense deployment of sensor nodes that collectively communicate event information to the sink. However, due to the spatial correlation between sensor observations, it is not necessary for every node to transmit its data. Exploiting spatial correlation in the context of collaborative nature of the WSN brings significant potential advantages
for the performance enhancement of communication protocols in WSN. Intuitively, data from spatially separated sensors is more useful to the sink than highly correlated data from closely located sensors. The suppression of redundant transmissions is achieved by collaborative regulation of medium access. A sensor node can act as a representative of several other nodes observing correlated data. CMAC protocol is designed with two components: Event MAC (E-MAC) and Network MAC (N-MAC). E-MAC filters out the correlation in sensor records while N-MAC prioritizes the transmission of route-thru packets.

Related Work
  • Vuran, M.C. and Akyildiz, I.F., "Spatial Correlation-based Collaborative Medium Access in Wireless Sensor Networks," Submitted for Publication, July 2004.
Back to Top

Data Link Protocol: Error Control

Traditionally, error control has been viewed as a means to improve energy efficiency at the cost of bandwidth efficiency. In sensor nets, they are better described as techniques to improve performance at the cost of reduced lifetime. Additional energy consumption in transmitting parity bits and encoding/decoding energies must be taken into account in evaluating the coding gain in sensor nets. From this perspective, coding may not always be energy efficient. This is one of the areas of our current research.

Related Work

  • Sankarasubramaniam, Y., Akyildiz, I.F., and McLaughlin, S.W., "Energy Efficiency based Packet Size Optimization in Wireless Sensor Networks," in Proc. First IEEE International Workshop on Sensor Networks Protocols and Applications (SNPA'03), Anchorage, Alaska, USA, May 2003 (held in conjunction with ICC'03).  
Back to Top

Power Management Scheme

Sensor nodes are very sensitive to power, because their operational lifetime are related to their battery live.  They are non-functional if they run out of battery, and they are usually deployed in a sensor field with a small battery.  At some scenarios, such as battle and toxic zones, replacing the batteries of these sensor nodes are impossible.  By managing the power usage of these sensor nodes, the lifetime of the whole sensor network can be extended.  The power management scheme integrates with the sensor node's application, routing protocol, MAC protocol, and physical layer.

Back to Top

Time Synchronization Scheme

Small intelligent devices can be deployed in homes, plantations, oceans, rivers, streets, and highways to monitor the environment.  Events such as target tracking, speed estimating, and ocean current monitoring require the knowledge of time between sensor nodes that detect the events.  In addition, sensor nodes may have to time-stamp data packets for security reasons.  With time synchronization, voice and video data from different sensor nodes can be fused and displayed in a meaningful way at the sink.  Instead of time synchronization between just the sender and receiver during an application like in the Internet, the sensor nodes in the sensor field have to maintain a similar time within a certain tolerance throughout the lifetime of the network.  Combining with the criteria that sensor nodes have to be energy efficient, low-cost, and small in multi-hop environment, this requirement makes an interesting problem to solve. 

Related Work

  • Su, W. and Akyildiz, I. F., "Time-Diffusion Synchronization Protocol for Sensor Networks," to appear in IEEE/ACM Transactions on Networking, Feb. 2005.
Back to Top

Localization Scheme

As the state-of-the-arts use signal sources to self-calibrate, there is a need to have localization methods that do no require signal sources. This is essential because signal sources may die or may be out-of-range.  In addition, methods based on beacons may not be deployed in different environments such as underground caves, and low-end sensor units may behave non-linearly when determining the range between sensor nodes. As a result, sensor nodes should collaboratively work together to provide a better estimate of the relative positions. If each node knows the relative positions of its neighbors, the location of any event can be determined by aggregating the relative positions of the nodes along the route.

Back to Top

Physical Layer

Physical layer research is driven by power-aware modulation and hardware design rather than targeting high data rates as in most other communication systems. Binary modulation techniques have been shown to be more energy efficient under start-up dominant conditions encountered in low-power short range wireless transceivers. Adaptive transmit power and dynamic voltage scaling are a couple of energy-efficient hardware strategies for sensor nets.  

Back to Top