The primary objective of the International Chair IoTAD-CEO led by Tadashi Matsumoto is to consider Distributed Multi-terminal Lossy Source Coding with applications to Beyond 5G and 6G Wireless Communications systems. In this paradigm, several terminals send their observations to a fusion center, and the original observations should not necessarily be losslessly reproduced at the receiver side. For instance, in Internet-of-Things (IoT) framework, the purpose is NOT to reproduce the data losslessly but to make a decision based on the lossy (distorted) data sequences collected by the receiver. The distortion may happen in the data gathering process (or equivalently observation process), and/or because of lossy encoding for compression of which encoded sequence may be noise-corrupted in the transmission channels. In this case, unlike in conventional approaches, the requirement is to satisfy an error probability requirement on the final decision.
Examples of applications:
Theoretical basis of this problem is Chief Executive Officer (CEO) problem, followed by decision making using Hypothesis testing framework. Therefore, the project intends to combine knowledge from both information theory and Machine Learning fields.