Wireless SensorActuator Networks for Industry Control Wireless SensorActuator

  • Slides: 1
Download presentation
Wireless Sensor-Actuator Networks for Industry Control Wireless Sensor-Actuator Networks (WSAN) comprises of sensors and

Wireless Sensor-Actuator Networks for Industry Control Wireless Sensor-Actuator Networks (WSAN) comprises of sensors and actuators (or actors) where sensors are low processing-complexity, low energy disposable devices used for sensing and collecting data while actors make decisions and take action based on sensed data, have higher processing capacity and are less energy constrained than the sensors. The requirements imposed on WSANs are varied and depend on the applications. Considering industry control and automation, Wireless Sensor-Actuator Networks for Industry Control (WSANIC) should have high survivability and capability to support data, event and task prioritization. Security is an important factor to take into account. These requirements are best addressed at the network architecture. Lack of adequate investigation of architectural aspects can result in development of suboptimal protocols and solutions. Intuitively, the first step then in the design and development of an efficient WSANIC should be an architecture that addresses survivability and several other constraints encountered in an industry. Based on this architecture protocols to address the communications requirements should be developed. The architecture shown in figure above was developed keeping the industry constraints in mind. We then developed an integrated medium access and routing protocol at layer 2 that is suited to the architecture and addresses the WSANIC communication requirements. The medium access was interruptible and used a loosely scheduled neighbor knowledge based approach to reduce collisions. Exponential backoff typical in random access protocols was avoided and instead a logical turn taking approach based on the neighbor knowledge was used. The neighbor knowledge was also used to implement the multi-meshed tree routing algorithm, with trees originating at the ACPs and culminating at the sensors and actors, providing overlapping coverage and route robustness. Year 2006 -2007, Funded IBM SUR grant, Nirmala Shenoy, Matt Cao