JOURNAL OF COMMUNICATIONS (JCM)

ISSN : 1796-2021

Volume : 3 Issue : 5 Date : October 2008

**Cell Site Power Characterization for Multi-rate Wireless CDMA Data Networks Using Lognormal Approximation**

Ashraf S. Hasan Mahmoud

Page(s): 47-53

Full Text: PDF (407 KB)

**Abstract**

An analytical model is developed to solve the power-rate assignment problem for multi-rate CDMA systems and

calculate the probability density function (PDF) for the downlink traffic power as a function of the bit rate assignment

scheme, the target link quality figure, and the other network parameters. This paper proposes the use of the

lognormal random variable as an approximation for an empirical parameter (denoted herein by fi) that is needed to

solve the energy per-bit to noise spectral power density equation for scheduled data bursts. The approximation

depends on matching the mean and standard deviation of the lognormal variable to the mean and standard deviation

of the empirical parameter fi where the matching depends only the radio frequency (RF) path loss model and the

shadowing process standard deviation. The paper presents example calculations for the cumulative probability

distribution (CDF) of the downlink traffic power and outage probabilities for a general CDMA system assuming

specific scenarios of data bit rate assignments. The study also compares between analytical results and those

obtained using conventional Monte-Carlo simulations. Comparisons indicate that the developed model provides an

appropriate mechanism to approximate both the PDF or CDF, and the outage probability for the cell site total

downlink traffic power.

**Index Terms**

code division multiple access, radio resource management, call admission control, burst scheduling

ISSN : 1796-2021

Volume : 3 Issue : 5 Date : October 2008

Page(s): 47-53

Full Text: PDF (407 KB)

calculate the probability density function (PDF) for the downlink traffic power as a function of the bit rate assignment

scheme, the target link quality figure, and the other network parameters. This paper proposes the use of the

lognormal random variable as an approximation for an empirical parameter (denoted herein by fi) that is needed to

solve the energy per-bit to noise spectral power density equation for scheduled data bursts. The approximation

depends on matching the mean and standard deviation of the lognormal variable to the mean and standard deviation

of the empirical parameter fi where the matching depends only the radio frequency (RF) path loss model and the

shadowing process standard deviation. The paper presents example calculations for the cumulative probability

distribution (CDF) of the downlink traffic power and outage probabilities for a general CDMA system assuming

specific scenarios of data bit rate assignments. The study also compares between analytical results and those

obtained using conventional Monte-Carlo simulations. Comparisons indicate that the developed model provides an

appropriate mechanism to approximate both the PDF or CDF, and the outage probability for the cell site total

downlink traffic power.