Ultra-fast lasers and photodiodes with bandwidths up to Ka-band, which were only available in research laboratories until recently, are now available commercially. New technologies and applications are being realized with these devices. A short list of these subjects might include high-speed communication systems, antenna remoting, radar, high speed computer interfaces, sensors, personal communication networks, phased array antenna feeds and a more recent topic, optically controlled semiconductor devices. Although the relatively high cost and low availability of high-speed optical devices still prohibit widespread usage, adequate design, test and characterization techniques for these bandwidths must still be developed.
Furthermore, extending F.O. link operating bandwidths to millimeter-wave frequencies requires an different design approach to that of microwave links. At millimeter wave frequencies, electrical circuit behavior and the performance of some optical devices are fundamentally different. Electrical parasitics, transitions and circuit radiation need to be considered in the design of transmitter and receiver modules. Additionally, the device's optical responses must be tailored to the new frequencies.
Currently, there are two major limitations in achieving millimeter-wave F.O. communication. One is the lack of laser diodes with sufficiently high modulation bandwidths and low noise characteristics. The other is the lack of external modulators with wide bandwidths, low drive voltages and low optical loss. Direct modulation of laser diodes is in practice, limited to approximately 25 GHz. Laser diodes are also capable of generating harmonic output which in effect doubles or triples the available bandwidth however, this compromises the stability, reliability, and linearity of the device. In the case of external optical modulators (EOM), typically LiNbO3 based Mach-Zehnder interferometers offer the best overall high-speed performance and manufacturing ease. Unfortunately, broadband operation is limited by conductor loss, velocity mismatch between lightwave and electromagnetic waves, and the existence of surface modes restricting the current ceiling of operation to about 50 GHz. Photodiode technology has matured well beyond the modulated sources and presents few obstacles to achieving high-speed F.O. link operation.
To extend F.O. link bandwidths beyond present limitations and to provide optimal performance from available devices, numerous system architectures and lightwave modulation techniques have been proposed and analyzed. Methods of improved light coupling, reactive impedance matching to devices, custom packaging, and use of optical fiber amplifiers have also proven to be effective in improving the performance of F.O. links. Nearly all of the methods tend to compromise performance with link simplicity. In most instances, the particular application determines the acceptable trade-offs.
At ATR, the ultimate goal is to develop the technology which would be used in personal communications systems. Such a concept is illustrated with the example of a wristwatch radio telephone which would provide instant access to any other terminal world-wide. Although there are numerous technologies to be developed such as packaging, battery, display etc., the one problem which remains is the delivery of the data signals to each of these terminals.
We envision a network based upon current cellular telephone communication systems. However, the cell diameter would be much smaller, perhaps several hundred meters or less. This small diameter arises from the limited transmission power of such a portable terminals. Thus within an installation of such a system, the number of cells increases quite rapidly. This presents a problem not only in power distribution, but also bandwidth. Future systems will most certainly require bandwidths exceeding the capacity of any copper type of installation. Thus fiber optic (F.O.) links are the only viable solution to providing lossless transmission with wideband capability.
Additionally, the move towards microwave and millimeter-wave frequencies presents many challenges to the F.O. link designer. The work which I have undertaken at ATR hopefully addresses some of the needs for these applications and provides several unique solutions. In the most general survey, it was found that to extend the carrier frequencies of the fiber optic links from the current microwave region upwards into the millimeter-wave bands, external modulation techniques provide the most promising cost/performance ratio, as well as lend themselves to currently available technology. The results of this survey are presented in the SPIE paper entitled Fiber optic link architectural comparison for millimeter-wave transmission, published in 1992.