| Summary: | The application of the nonredundant error correction (NEC) technique to the new North American and Japanese Digital Cellular modulation standard, the π/4-shift DQPSK modulation format, is proposed, analyzed and evaluated. Due to the nature of the mobile cellular communication
channel, it is assumed that the π/4-shift DQPSK system is operated in a combined additive white Gaussian noise (AWGN) and cochannel interference (CCI) environment as well as in a frequency-nonselective fading environment.
The NEC techniques can be accommodated to the π/4-shift DQPSK by a modification of the NEC receivers for the DQPSK following which the performance of the NEC receivers with single-, double- and triple-error correction capability are theoretically analyzed and evaluated. The most elaborate system analyzed is the triple-error NEC receiver which employs four differential detectors with delay elements of one up to four symbol duration long and which requires the computation of 12 syndromes for the correction of error symbols.
For the CCI, the general model which includes M statistical independent interferers also employing the π/4-shift DQPSK modulation format is adopted. The theoretical symbol error rate (SER) versus carrier-to-noise (C/N) ratio have been obtained having M and the carrier-to-interference (C/I) ratio as parameters. These performance evaluation results indicate significant performance improvements over conventional differentially detected π/4-shift DQPSK systems without requiring any bandwidth expansion. For example, at a SER = 10ˉ⁴ and for C/I = 14 dB and M = 6, gains of more than 7 dB have been obtained. Compared with a coherent π/4-shift QPSK system operated in the same environment, this triple error correcting NEC is inferior by only 1.5 dB. Some of these theoretical results have also been verified by computer simulation. The gains offered by the NEC receivers have been found to increase as C/I decreases and/or M increases. In addition to the performance improvements, significant error floor reductions (of at least one order of magnitude) have been observed.
For the fading channel, the theoretical error rate equation for the single-error correcting NEC receiver is newly derived. Since numerical evaluation of the derived equation is extremely time consuming, computer simulations were used to obtain the performance evaluation result of π/4-shift DQPSK system employing single- and double-error correcting NEC receivers. In general, the improvement are not as high as that in the CCI environment. For example, at a BER = IOˉ² and for a Rician fading channel with the K-factor of 1 dB with BDT = 6.29, a performance gain of 6 dB is achieved. The gains offered by the NEC receivers increases as the K-factor decreases and/or when the BDT is large.
Since the NEC technique does not require any bandwidth or signal constellation expansion as do other coding schemes, it is a powerful and attractive technique to increase the capacity of digital communication systems operated in CCI controlled (frequency reuse) environment, such as the new all digital North American and Japanese mobile/cellular network. The significant improvement of the NEC receivers in a very fast fading environment suggests the NEC receivers can also be applied to communication applications in which the speed of the mobile unit is very high, for example, in aeronautical communication systems. Finally, it is noteworthy that the results obtained in this thesis for the π/4-shift DQPSK systems are directly applicable to the DQPSK systems in a linear channel. === Applied Science, Faculty of === Electrical and Computer Engineering, Department of === Graduate
|
|---|
