Abstract
Phased arrays in 5G millimeter wave (mmW) communication are used at both link ends to separate users and to enhance the link range. The radio hardware’s RF performance metrics vary due to part-to-part variations within component tolerances. In this paper, we derive probability density functions (PDFs) of over-the-air (OTA) transmission (TX) powers for individual transmitter (Tx) paths and total TX power of the 5G mmW array using Monte Carlo simulations based on component tolerances. The PDF of the TX power individual path of the array follows a log-normal distribution on a linear scale. Hence, the total array’s OTA TX power is a sum of independent linear scale log-normal powers when the power combination happens in OTA towards the main beam direction. We show that the Tx array’s mean power increases when multiple log-normally variated individual TX powers are summed together. Additionally, we prove that a link budget maps to a link range in a non-linear way, skewing the normally distributed link budget to a log-normally distributed link range. The link ranges based on RF system parameter analysis, RF system calculation, and OTA measured system error vector magnitude (EVM) based link range estimation had an excellent match when a 16-QAM 5G NR signal was used. The calculated link ranges are 410 and 434 meters, while the OTA measured system EVM measurement-based link range is 415 meters. For example, uncalibrated Tx array paths would increase the mean of the array’s TX power by 1.2 dB, and the link range increases from 415 to 498 meters, but simultaneously the standard deviation increases from 26 to 51 meters.
Highlights
If the antenna gain variation with σ = 0.59 dB and the consecutively measured TX variation with σ = 3.11 dB are combined to the individual TX path OTA variation σ = 3.16 dB, the TX EIRP is increased to 41.2 dBm with a standard deviation of 0.89 dB
Statistical information is needed for quality assurance purposes during the product development and manufacturing phases
We showed that individual path TX powers’ mean and variance affect the increase of the EIRP of the array compared with the EIPR calculated using average values of individual TXs
Summary
If the standard defines only one limit, the device manufacturer can set its design target based on the estimated or the measured variation of prototypes or products. Accurate quality level estimation requires that the probability density function (PDF) of the radio parameter is known either based on simulations and measurements of prototypes or, preferably, based on a set of measurements of mass production samples. A 5G mmW radio link range estimation based on short-range measurements is presented in [12].
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
