The linear minimum-mean-square error (LMMSE) estimator is herein derived to estimate phase-noise of Levy statistics (including Wiener phase-noise) and of arbitrarily large magnitude. This estimator may be pre-set to any latency and any number of tabs. This estimator depends only on the signal-to-(additive)-noise ratio and the phase-noise variance, but requires no matrix-inversion.

When a signal is highly saturated and clipped by an analog-to-digital converter, the maximum-likelihood power estimator for the clipped signal is derived analytically for Gaussian-distributed input signal. The estimator error is also found numerically by simulation. The power estimator may find its applications in automatic gain control.

Chromatic dispersion induces to a phase-modulated optical signal a constant phase rotation that was never included in most previous studies. When the constant phase shift is removed by a phase-locked loop, the dispersion tolerance of a coherent-detection system is increased. The dispersion tolerance of phase-shift keying (PSK) signal is compared with differential PSK (DPSK) signals by numerical simulation and experimental measurement. Contrary to conventional belief, PSK signal has larger dispersion tolerance than DPSK signal.

Nonlinear phase noise induced by the interaction of the fiber Kerr effect and amplifier noises is a quadratic function of the electric field. When the dependence between the additive Gaussian noise and the quadratic phase noise is taken into account, the error probability for differential phase-shift keying signals is derived analytically. Depending on the number of fiber spans, the signal-to-noise ratio penalty is increased by up to 0.23 dB, due to the dependence between the Gaussian noise and the quadratic phase noise.

The well-known Bahl-Cocke-Jelinek-Raviv (BCJR) algorithm is used as a minimum mean-squared error (MMSE) decoder for combined trellis-coded quantization and modulation. When the BCJR algorithm is used as decoder, the optimal trellis-coded encoder is derived. Both trellis-coded quantizer and the soft-decoder can be optimized iteratively for minimum overall distortion from transmitter input to receiver output. The optimal encoder/decoder is simulated for Rayleigh fading channel. Significant performance improvement is achieved for both memoryless Gaussian and uniform source, especially for very noisy channel.

The maximum a posteriori (MAP) algorithm is used as a minimum mean-squared-error decoder for combined trellis-coded quantization and modulation. The optimal trellis-coded quantizer is derived when the soft-decoding MAP algorithm is used as a decoder. The trellis-coded quantizer and the soft decoder are optimized iteratively for minimum overall distortion from transmitter input to receiver output. Significant performance improvement is achieved for both memoryless Gaussian and uniform source, especially for very noisy channels.

High-capacity multiwavelength ring networks with bidirectional WDM add/drop multiplexer (WADM) are analyzed and demonstrated. All channels can be added/dropped independently in each direction. The capacity of a bidirectional ring is found to be approximately twice that of a unidirectional ring. An eight-wavelength WADM is demonstrated for a data rate of 10 Gb/s per channel, providing an overall capacity of 80 Gb/s.

A very simple speech coder is investigated based on discrete-cosine transform (DCT) and run-length coding. The speech signal is first passed through a DCT, quantized by a uniform quantizer, and run-length coded to achieve good compression ratio. Not suitable for wireless communication systems, the speech coder is perfect for Internet and speech storage applications. With a performance far better than most speech coding for wireless applications, low data-rate speech of 4 kbit/s can be achieved with a signal-to-noise ratio (S/N) of about 11 dB, comparable to 8 kbit/s VSELP and CS-ACELP speech coder.

We present an algorithm for design of a joint source-channel coder using a channel-optimized quantizer and multicarrier modulation. By changing the power of each subchannel in the multicarrier modulation system, different degrees of error protection can be provided for different bits according to their importance. The algorithm converges to a locally optimum system design. Compared to a Lloyd-Max scalar quantizer or a LBG vector quantizer using single-channel transmission, our optimized code can yield substantial performance improvements. The performance improvements are most pronounced at low channel signal-to-noise ratios.

We have proposed and analyzed a combined source-channel coding scheme using multicarrier modulation. By changing the power and modulation of subchannels carrying different bits of the quantized signal, the channel-induced distortion can be minimized. We derive an algorithm for the allocation of power among subchannels. Analog sources quantized by scalar and vector quantizers are considered. For a Gaussian source scalar-quantized to 8 bit, multicarrier transmission yields a 13 dB improvement of signal-to-distortion ratio over a single-carrier system. Quantizers having a smaller number of bits yield smaller improvements. We consider both full-search and binary tree-search vector quantizers using a natural binary coding scheme. For a vector quantizer using 2 b/sample, multicarrier transmission yields an improvement in signal-to-distortion ratio that lies between 0.3 and 0.8 dB, depending on the quantizer design.

For a clipped channel with a Gaussian input, a predistortion/restoration technique to reduce clipping-induced nonlinear distortion is proposed and analyzed. The input signal is processed by a nonlinear predistorter circuit, reducing the probability of clipping. The receiver output signal passes through a restorer having an inverse transfer characteristic, which yields the original signal. For both one-sided and two-sided limiter channels, the optimal predistortion curves are determined analytically. A limiter channel with Gaussian input may be used to model clipping-induced nonlinear distortion in optical-fiber common antenna television (CATV) distribution systems using multiple intensity-modulated subcarriers. When applied to an amplitude-modulated vestigial-sideband (AM-VSB) CATV system, the optimal predistortion curves yield sensitivity improvements of 5.3 and 4.4 dB for oneand two-sided limiter channels, respectively.