Policies for Sensor Planning

SANY D2.3.4 Specification of the Sensor Service Architecture V3 Doc.V3.1 Copyright © 2007-2009 SANY Consortium Page 188 of 233 integrity of the measurement data. The more complex form of monitoring consists if there are multiple sensors at the same time. An example is the video monitoring of a measurement station in order to detect external factors that might affect the measurement process. Other monitoring scenarios could involve sensor service degradation. The main limitation in scenarios involving mobile sensor networks is the power supply usually batteries of each individual network node. Depending on the network technology and geographic distribution of the nodes a failure of one node might render the entire network unusable. In these cases a vital monitoring aspect is the node residual power. Based on this information a sensor network administrator can make management decisions. Examples of such decisions include selection of network optimisation strategies e.g. adjusting the network routing policies or node reporting sampling frequencies in order to reduce the power consumption of individual nodes or entire network. In the later scenario a good indication of the residual power of a node is the power supply voltage. On a conceptual level as described in section 6.6.2 in the SensorSA Management Architecture this information constitutes the actual observation data and should be handled the same way as any other observation coming from the sensor e.g. ambient air temperature or humidity. Therefore, the observation data about power supply voltage shall be encoded as an observation according to the OGC Observation and Measurement model Cox, 2007. Consequently, the observation data can be stored and accessed by means of an OGC Sensor Observation Service see section 8.2.2. Additionally the Sensor Alert Service see section 8.2.4 can be used to define event conditions and generate notifications about the observation data, e.g. to send an e-mail notification to the sensor network administrator whenever a node battery voltage drops under a defined threshold.

10.4. Policies for Sensor Planning

Examples of sensor planning tasks in a sensor network include sensor configuration, sensor calibration or the actual initiation of a measurement. Whenever a measurement is triggered or prepared, the sensors involved must be configured for the specific measurement or measurement series. This can be achieved by the Sensor Planning Service SPS as described in section 8.2.3. Although the same operation submit is invoked for both planning and configuration the slight difference is the observation response. For planning the response encompasses observation data whereas the result returned upon configuration will contain the success status of the configuration step. One obvious advantage is the possibility of planning configuration tasks. In general, sensor planning includes different interaction models or patterns. Some sensors allow synchronous interaction patterns, i.e. the service responds directly to incoming requests. An example would be an instance of an SPS that provides a facade for a simple forecasting model. This service, at least theoretically, could start unlimited parallel processes. Concurrent users don‟t compete for limited resources and the service can report the successful execution of the requested tasking right away. Other sensors require asynchronous interaction patterns. This is the case if multiple users have to share a limited resource and the execution of the tasking cannot be handled instantaneously. An example would be a satellite that could at any moment in time observe a single scene only. If this satellite is equipped with an optical sensor, the observation depends, SANY D2.3.4 Specification of the Sensor Service Architecture V3 Doc.V3.1 Copyright © 2007-2009 SANY Consortium Page 189 of 233 among other factors, on the cloud coverage. Thus, the tasking request might consume any amount of time before being fully executed.

10.5. Policies for Access Control