Fred C. Kellerman

Serial concatenation schemes for PSK waveforms vs. turbo codes

For radio communication systems, powerful error correction codes are necessary to operate in noisy and fading channel conditions. Iterative forward error correction schemes like Turbo codes can achieve near Shannon capacity performance on memory-less channels and also perform well on correlated fading channels. The key to the excellent decoding performance of the Turbo coding systems is the BCJR algorithm in conjunction with the iterative processing of soft information. A very popular modulation technique is Differential Phase Shift Key (DPSK) which is not only a simple non-coherent modulation and demodulation technique; it is also a recursive rate one code. Combining DPSK with a single convolutional code structure as an iterative inner outer forward error correction system can provide excellent Turbo like performance. Bit Interleaved Coded Modulation with Iterative Demodulation (BICM-ID), another powerful iterative technique achieves near Turbo code performance with significantly less mips. We will also introduce and compare with the latter systems yet another novel iterative scheme that utilizes coherent demodulation in conjunction with convolutional codes. This new system can easily be extended to higher order modulations such as 16 and 64 Quadrature Amplitude Modulation (QAM) while only requiring modest amounts of processing power. Monte Carlo simulation results will be shown for the Additive White Gaussian Noise (AWGN) channels.

Performance of concatenated convolutional codes: coherent vs. non-coherent

For radio communication systems, powerful error correction codes are necessary to operate in noisy and fading channel conditions. Iterative forward error correction schemes like Turbo codes can achieve near Shannon capacity performance on memory-less channels and also perform well on correlated fading channels. The key to the excellent decoding performance of the Turbo coding systems is the BCJR algorithm in conjunction with the iterative processing of soft information. A very popular modulation technique is Differential Phase Shift Key (DPSK) which is not only a simple non-coherent modulation and demodulation technique; it is also a recursive rate one code. Combining DPSK with a single convolutional code structure as an iterative inner outer forward error correction system can provide excellent Turbo like performance. Bit Interleaved Coded Modulation with Iterative Demodulation (BICM-ID), which is a similar iterative coding system that allows full coherent processing, will be analyzed and compared to the DPSK BCJR iterative system. Monte Carlo simulation results will be shown for the Additive White Gaussian Noise (AWGN) and Rayleigh fading channels.

Performance of concatenated convolutional codes with m-ary differential phase shift key modulation

For radio communication systems powerful error correction codes are necessary to operate in noisy and fading channel conditions. Iterative forward error correction schemes like Turbo codes can achieve near Shannon capacity performance on memory-less channels and also perform well on correlated fading channels. The key to the excellent decoding performance of the Turbo coding systems is the BCJR algorithm in conjunction with the iterative processing of the soft decision information. A very popular modulation technique is Differential Phase Shift Key (DPSK) which is not only a simple non-coherent modulation and demodulation technique, it is also a recursive rate one code. Combining DPSK with a single convolutional code structure as an iterative inner outer forward error correction system can provide excellent Turbo like performance. Monte Carlo simulation results will be shown for the Additive White Gaussian Noise (AWGN) and Rayleigh fading channels for 1, 2, 3 and 4 bits per symbol DPSK.

Some practicalities of popular error correcting codes on fading channels

For real world communication systems that operate in correlated fading channels, perfect channel state information is not always available. In the literature, performance curves for error correction codes are usually plotted from either closed form equations or simulations which assume perfect channel state information. While these methods of measuring the capabilities of error correcting codes do serve a theoretical purpose, they do not necessarily demonstrate how well a code will perform under non-ideal conditions. The goal of this paper will be to compare LDPC and Turbo codes and determine how well they perform when perfect channel state information is not available at the receiver. Bit error rate and some block error rate performances will be provided for the AWGN, Rayleigh and 1Hz 1ms multipath fading channel. The results of this paper may provide communication systems designers some useful insight into the actual performance of error correcting codes in real-world scenarios.

Differential space-frequency coding for a multipath fading channel

This paper will investigate differential space frequency coding and its applicability to multipath fading High Frequency (HF) radio channels. Orthogonal Frequency Division Multiplexing (OFDM) will be combined with differential Alamouti space frequency codes to measure performance on the Watterson HF channel model. Differential coding facilitates non-coherent reception and can thus also reduce receiver complexity. Numerical results will be shown for the CCIR poor (2ms, 1Hz) and extra poor (2ms, 2Hz) channel conditions for a system comprised of 2 transmit antennas and a single receive antenna. For comparison, performance will also be shown vs. a new single transmit and receive antenna HF OFDM Code Division Multiple Access (CDMA) scheme.