Radio Propagation Models for Pico-cellular Networks
Authors:Melissa Johnson, Francois Quitin, Dinesh Ramasamy
Mentor:Upamanyu Madhow, Professor of ECE, University of California Santa Barbara
Mobile phone networks use macro-cell base stations in order to provide coverage capabilities for large areas. Pico-cells, which have small coverage areas have been proposed to increase network capacity in urban scenarios. One concern when deploying pico-cells is their poor base station antenna placements, making radio propagation different from that of traditional macro-cells. Understanding radio propagation for pico-cellular deployments is essential for interference management. Knowing whether macro-cells and pico-cells can be used to supplement each other when being co-deployed is an important consideration for designing such networks.
We undertook a measurement campaign to evaluate different path loss models, which predict the average power decay as a function of distance for pico-cellular networks. The measurements were performed with software-defined radios as base station transmitters and mobile receivers. We drew two conclusions from our measurements.
The first conclusion is regarding path loss models. For traditional macro-cells, path loss is modeled by a linear path loss model. Our measurements indicate that, for pico-cells, the Franceschetti model which allows an extra exponential decay in received power in addition to the inverse square law is more accurate. The distance beyond which power loss in this model is much faster than that predicted by the inverse square law (known as the breakpoint distance) is in the 100-150m range. This indicates that pico-cells with a 150m cell radius will provide good coverage to their cell without considerable interference to neighboring cells.
The second conclusion is regarding co-deployment. As the mobile station moves around the deployment region, the signal strength received from the macro-cell base station and the pico-cell base station were measured simultaneously. The correlation between these two signals was found to be low, suggesting that macro-cell base stations can take over service when the channel between the mobile receiver and the pico-cell base station is weak.