Message Identification for Task-Oriented Communications: Exploiting an Exponential Increase in the Number of Connected Devices.

The increase in the amount of data sent through the Internet is expected to scale up and accelerate in the coming years, along with the increasing number of connected devices. Current data rates via optical fiber communication channels are approaching the channel capacity and eventually, wireless communication will follow a similar trend. This means that message transmission may not be able to keep up with traffic demand in the future. Optimizing communication tasks and reducing the needed number of utilized radio resources will thus play a crucial role in goal-oriented communications. The latest theoretical and experimental works on message identification, a post-Shannon communication scheme, show promising results for addressing this problem. Under this scheme, the objective at the receiver is no longer to retrieve the content of the transmitted message, but to verify whether the sender is transmitting a specific message, in which the receiver may be interested or not. The sender does not know in advance which message the receiver is interested in. By using local randomness at the encoder, this allows for an exponential gain in the number of messages in identification compared to the available messages in classical transmission, potentially saving radio resources. In this paper, we review the pros and cons of identification and discuss its potential applications for task-oriented communications.