Review of 'Automata Theory: An Engineering Approach' (Aleksander, I., and Hanna, F. K.; 1975)

including the spectral density. This last clearly reveals the inevitable noise level originating in the discrete,nature of the detector model. This chapter also examines the relation between the statistics of the shot-noise process and those of the underlying intensitJr process. The concluding section of the chapter mentions some possible generalizations of the photodetector model; e.g., random electron gain. Chapters 2-4 comprise a thorough and quite readable treatment of the photodetection process for the communications engineer. However, there,are some gaps in the coverage. Counting statistics for non-Gaussian (e.g., lognormal) fields are only mentioned. More important, there is no discussion of the powerful techniques that have surfaced in recent years for the characterization and analysis bf random point processes, especially doubly-stochastic Poisson processes. These techniques, which are concerned primarily with the temporal evolution of counting processes rather #than with fixed-interval counting statistics, provide useful alternative detector models. Chapters 5 and 6 are devoted to the introduction of direct-detection and heterodyne optical communication systems. The emphasis is on accurate waveform transmission; consideration of digital transmission, where symbol errors must be minimized, is deferred to later chapters. Performance is measured in terms of signal-to-noise ratio, with the exception of section 5.5 where the minimum mean squared error linear processor is discussed. A section in chapter 6 deals with the alignment and field matching problem for heterodyne systems. This is an important topic which is ignored in most introductory treatments of het-erodyning. Another section discusses multimode heterodyning. The chapter concludes with an examination of the effects of atmospheric turbulence on heterodyne systems. Digital systems are covered in detail in chapters 7 (binary systems) and 8 (block coded systems). Except for one section, consideration is restricted to the direct-detection format. In the heterodyne format, the reasonable assumption of a strong local oscillator allows the detected signal to be modeled as a known function in additive Gaussian noise, thereby yielding an equivalent problem with well-known solutions which the authors apparently saw no point in reproducing. Chapter 7 begins with discussions of binary transmission and maximum a posterior-i detection, and develops the likelihood ratio test along the usual lines. The remainder of the chapter applies the test to various situations to obtain receiver structures. In most cases, performance is evaluated in terms of error probabilities which are presented in graphical as well as equation form. There is also an interesting and useful section that deals with the effects …