HP INCFollow


We propose a hybrid, long-lived network connection mechanism that alternates

between polling and persistent operation modes based on the activity profile of

network-enabled printing devices. Our Machine Learning method is able to predict

“printing peaks”, proactively alternating the connection mode back and forth from

polling (for those moments of low activity) to persistent model (for high demanding

usage scenarios). This strategy combines the advantages of low consumption, low cost

polling connections, and at the same time provides low response times offered by

persistent connections. Our solution minimizes costs and maximizes network usage and

performance, respecting the availability requirement thanks to its prediction capability.

Besides the fact that our solution was thought for printers, it can also be applied for other

devices that require optimized connection management.

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Share Alike 4.0 License.