| Signal processing for RF communications and radar is experiencing enormous growth, as emerging applications require throughput in excess of {dollar}10sp{lcub}12{rcub}{dollar} oper/sec. Superior bandwidth and parallelism of photonics presents a number of insertion opportunities for such antenna systems. In time domain, of specific interest are true-time-delay and adaptive beamformers based on optical delays, as well as 2-D signal processing. The research objective has been to examine performance and accuracy-limiting issues of the aforementioned systems and propose novel approaches to enhance them.; First, analog adaptive processor with signal channelization scheme is studied and optimized. It is shown that 50 dB interference suppression over bandwidth 20 MHz is possible with a 20 ... 25 channels. Also, the intrinsic tradeoff between cancellation ratio and convergence speed has been identified, and switched-stepsize algorithm is proposed to mitigate it. Also, processor enhancement with single crystal fiber array (SCFA) has been sought, and SCFA fabrication issues are discussed in-depth.; Accuracy of optical matrix-vector processor (OMVP) for digital beamforming is related to the component quality and OMVP dimensions. It is shown that OMVP exhibits 6 ... 8 bit accuracy. The accuracy assessment is intended for OMVP design.; The problem of synthesizing fiber Bragg gratings (FBGs) for WDM delay lines is linked to the results of window theory. Performance of tapered gratings, including reflectivity, sidelobe level, bandwidth, and group-delay ripple, is quantified.; Finally, potential for embedding error control into 2-D optical correlators is discussed, and existence of coding gain is demonstrated. |
