Approaching fundamental limits on signal detection

Progress in the design of high performance photodetectors has been reviewed. On the basis of a general fluctuation theory, seemingly disjoint activities in this area are demonstrated to be of common origin. The conventional wisdom of designing low-noise photodetectors by using single species multiplication processes is compared with new approaches. Also, the noninstantaneous nature of the response to a photon is demonstrated to be not only useful but leading to near ideal amplification. The interplay between the evolution of new concepts, material limitations and performance requirements is elucidated. >