High‐rate, power‐efficient coding for optical communication: Design and performance evaluation

Abstract In the direct detection optical communication system which detects the optical pulse by counting the photons, it is known that the information transmitted by each photon has an important role in the performance evaluation of the error‐correcting code, in addition to the code rate and the decoding error rate. Especially, for the code which can achieve arbitrarily small error rate by increasing the code length, the performance of the code can be evaluated based on the code rate and the information efficiency. However, these two evaluation measures are in a trade‐off relation; and there is a limit if one attempts to improve both. From such a viewpoint, this paper determines the bounds for the code rate and the information efficiency and determines the limit of performance when the ideal error‐correcting code is applied to the direct detection optical communication. Then, as a code which can arbitrarily achieve a small error rate, a new code based on the iterated code is proposed. It is shown that if the code rate is less than 0.75, the proposed code can achieve a better information efficiency than the conventional code with the same code rate.