On a proposal for an optical receiver achieving high sensitivity (comment and reply)

The receiver structure introduced by K. Kikuchi (ibid., vol.4 p.195-7, Feb. 1992) is clarified. It is shown that although it improves the signal-to-noise ratio, it does not allow communication with fewer photons per bits than the traditional limit. K. Kikuchi responds to this letter and another one challenging the bit-error-rate performance of the receiver he proposed (ibid., vol.4, p.195-7, Feb.1992) and defends his results.<
>