Evaluation of Combined Availability Prediction Methods for Multicarrier Microwave Links

Nowadays, microwave backhaul can reliably provide several gigabits of capacity, in many cases due to the aggregation of multiple parallel links working at different frequencies. The combination of the E-band (80 GHz) and traditional microwave (MW) bands like 6 or 18 GHz is a commonly adopted solution in the industry. While the former band is used for high-density metropolitan applications by combining large capacity and high spectral efficiency at low spectrum costs, the latter is key to deployment in suburban areas. Multicarrier backhaul solutions are essential for backhaul evolution, especially to meet the capacity needs of 5G and beyond over the next 10 years [1]. While the availability prediction methods for a single carrier link are well-known and widely applied for link budget calculations, the method for estimating the availability of the multicarrier system's combined throughput (or simply the combined availability) needs to be developed. Current ITU recommendations have limited application in this latter case. Additionally, there exist software-based tools (like Pathloss 6) that use their own techniques to evaluate the outage relation between different carriers. The original research forming the basis for this article uses experimental data, obtained by performance monitoring of a multicarrier (dual-band or using 18 GHz and 80 GHz bands) test link over a period of 6 months, for actual combined availability estimation. The purpose of this article is to compare the obtained results with software simulation output based on ITU recommendations, as well as to evaluate the theoretical approach to combined availability calculation of a multicarrier link. The results showed a correlation between experimental and theoretical approaches, as well as between different algorithms of ITU-based theoretical approach, which will be discussed herein.