Date Added: Nov 2009
The authors study the energy - distortion tradeoff in communicating sources over noisy communication channels, such that the source can be reconstructed at the receiver within a target distortion. The 'Energy' refers to the cost per source sample of using the communication channel. The more energy that the transmitter has at its disposal, the more information it can transmit to the receiver. This potentially translates to lower distortion at the receiver. Thus, there is a fundamental energy-distortion tradeoff in this setting. They try to capture this tradeoff by defining a fundamental information-theoretic function E(D) which is the minimum energy required per source sample to achieve an average distortion D, for large number of source samples.