You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
It may be possible to save power by batching the data collection. Sensor nodes would collect a local historical batch of data samples, then when data is requested by the collector, the full batch would be sent together. This would enable a significantly longer collection period and radio sleep time.
This approach would also necessitate the logical separation of data sampling from data collection. This separation would open up some potential use cases where sampling time might be driven more by environmental conditions or would otherwise generally across nodes.
Some considerations to keep in mind:
How would receive buffer handling need to be different given the need to send larger payloads (ie. a batch of data samples vs a single data sample)?
What are some use cases that would benefit from the logical separation of sampling from collection?
Is there a worthwhile tradeoff? Does shutting down the RX cycle for long periods of time safe considerable power, or since we are effectively sending the same amount of data (just in batches instead of individual messages) does it all add up to about the same amount of power consumption?
What are some use cases that are not solved by this approach? E.g. any shutdown scheme precludes an environmental-state-driven alarm system as a potential use case. So this approach does not solve for that particular scenario, however separation of sampling and collection might be a helpful step in architecting for such a use case
The text was updated successfully, but these errors were encountered:
It may be possible to save power by batching the data collection. Sensor nodes would collect a local historical batch of data samples, then when data is requested by the collector, the full batch would be sent together. This would enable a significantly longer collection period and radio sleep time.
This approach would also necessitate the logical separation of data sampling from data collection. This separation would open up some potential use cases where sampling time might be driven more by environmental conditions or would otherwise generally across nodes.
Some considerations to keep in mind:
The text was updated successfully, but these errors were encountered: