International Journal of High Speed Electronics and Systems Vol. 16, No 2 (2006) pp. 559-566 © World Scientific Publishing Company FEASIBILITY OF AN OPTICAL FREQUENCY MODULATION SYSTEM FOR FREE-SPACE OPTICAL COMMUNICATIONS SERGE LURYI AND MIKHAIL GOUZMAN Department of Electrical and Computer Engineering, SUNY, Stony Brook, NY 11794-2350, USA We consider a free-space communication system based on optical frequency modulation (FM), where the information is encoded by a time-variable wavelength. As is well known, broadband FM systems use a transmission bandwidth that is larger than the signal’s information bandwidth, thus enabling an enhancement of the signal-to-noise ratio (SNR) and hence the effective information rate per unit transmitter power. Because of the atmospheric conditions, any optical free-space communication system, contemplated at a terrestrial level, must operate at mid-infrared wavelengths in the range λ = 2.5-2.8 µm. Development of rapidly tunable single-frequency lasers in this wavelength range is quite feasible, based on the current experience with tunable telecom lasers at 1.5 µm. Nevertheless, there is no currently available optical FM system. The main difficulty is associated not so much with the tunable optical sources, as with the implementation of a wavelength-discriminating receiver system that would take advantage of the enhanced SNR. In our view, the key enabling solution is optical superheterodyne with a local oscillator implemented as a tunable mid-infrared laser similar to that at the source. The intermediate frequency can be tuned to lie either in a frequency range directly accessible to electronic limiting amplifier and frequency discriminator or, in a multichannel system, to a second heterodyne in the terahertz range. 1. Introduction As is well known from radio-electronics, wideband frequency-modulation (FM) systems offer a trade of the bandwidth excess for signal to noise ratio, thus relaxing the transmitter power requirement as compared to AM transmission. Energy efficiency is essential for satellite communications, sensor networks and mobile platforms. The FM advantage is proportional to the squared ratio 2 ) / ( S f F ∆ of the range of frequency excursion ∆F to the signal bandwidth S f , see, e.g., a recent discussion by Hayes 1 . Thus current direct broadcast satellite systems are made possible by using a microwave bandwidth 28 = ∆F MHz to transmit each 3-MHz television channel, thus gaining nearly 20 dB in signal to noise ratio. The structure of a FM signal is illustrated in Fig. 1.