Analog diversity coding to provide transparent self-healing communication networks

pg 100 diversity coding, as introduced in Ayanoglu et al. (1990), is a method of protection against failures in a communication network or a storage system, which is based on introducing a digital error-correcting code across independent links. This technique makes efficient use of the extra network capacity needed for coding and has the additional advantages of being nearly instantaneous, not requiring a feedback channel, rerouting, or resynchronization. In high-speed (multi Gbps) networks, digital coding will be difficult to implement, and the purpose of the present paper is to demonstrate how diversity coding may be implemented in the analog domain using the discrete fourier transform (DFT). In particular, the authors show that the DFT is a continuous-amplitude maximum-distance separable code over the field of complex numbers when the transform kernel is a prime root of unity. This code can be used to generate self-healing or fault-tolerant communication networks for continuous- or discrete-amplitude signals, as long as continuous-amplitude parity channels are available. The authors describe electrical and optoelectronic implementations, and a signal estimation approach to combat channel noise and thereby improve the performance of the analog diversity coding system. The most important advantage of this technique is in greatly simplifying the encoders and decoders of diversity coding systems for high-speed networks, such as fiber-optic wavelength division multiplexed networks. Application of analog diversity coding to systems with analog sources, such as telemetry systems is also possible.< >