Propagation measurement-based probability of error predictions for digital land-mobile radio

This paper demonstrates that closed-form equations can be used to calculate error rates for /spl pi//4-shift differential quadrature phase-shift keying (DQPSK) directly from time-series measurements of impulse-response data on frequency-selective land-mobile radio channels. Results are quantitative and free from uncertainties that can accompany such predictions when they are made using qualitative relationships based on assumed channel characteristics. Calculations using the demonstrated method show that significantly greater signal-to-noise ratios are required on macrocellular (MC) channels than those required on microcellular (/spl mu/C) channels for the same performance, In addition, it is shown that MC channels require channel protection, even for IS54 performance levels, whereas higher data rates (up to 250 kbps at 10/sup -3/) and lower error probabilities (down to 10/sup -4/ at 48.6 kbps) are possible on line-of-sight (LOS) /spl mu/C channels without such protection. This is considered to be crucial information in the design of future systems. Through the analysis of quantitative results, not available elsewhere, it is shown that correlations between rms delay spread and performance measures, such as unprotected channel-error-rate floors and maximum data rates, are poor. Recommendations are, therefore, made regarding other parameters (the Rician K ratio and another parameter Vol, defined in the present paper) that can be derived through simple propagation measurements and used to predict these performance characteristics with greater assurance.