Network assessment of spectrum availability for mobile network in traffic intensive areas

Abstract

Recently 5th Generation(5G) network traffic is gradually increasing with the expansion of 5G service, and these trends of 5G traffic causes the needs for a continual spectrum supply. However, since frequency is a limited resource and currently the mid band above 3 GHz which is 5G main service band has a problem with short coverage, it is necessary to refarm the redundancy of low-band spectrum of existing mobile technology such as 4G into 5G. The quantization of this spectrum redundancy and availability can be achieved through network assessment. Meanwhile, network assessment should be estimated based on the traffic-intensive area in consideration of regional traffic density distribution since spectrum resources are not shared regionally. This dissertation proposed the methodology for network assessment using volume-based and spectrum-based evaluation methods with the traffic-intensive regions scale approach. First, this dissertation proposed the concept of a virtual sample region which is a logical region consisting of higher dense cells for spectrum utilization or traffic in a dense area. Then traffic trend using the correlation with national traffic is predicted in this region, estimate network capacity using spectral efficiency, bandwidth, and spectrum reuse factor (the number of base stations). Finally, the capacity margin of the network is derived from difference between traffic volume and network capacity. Secondary, virtual cell concept corresponding to the capacity of radio resources in virtual sample regions is proposed for spectrum-based network evaluation. In this dissertation the analytical domain is transformed from the usage rate to the usage count domain representing trends of traffic, bandwidth, and the number of cells in the network. Subsequently, the trends of radio resource block (RB) usage counts in the virtual cell is derived from the correlation between the RB usage counts and national traffic, and the cell-weighted RB usage rate is predicted by the ratio of it to the total number of RB. For the numerical evaluation of network assessment for availability, the forecast was undertaken by December 2028 using measured data of the top 10% of 4G traffic density cells in Seoul, Korea from March 2017 to March 2022. According to the proposed method in this dissertation, the capacity margin of a 4G network has a minimum value of 32,693 GB/h (33.8%) in Sep. 2018, and increases to 115,713GB/h (80.9%) by December 2028. And it is analyzed that the spectral availability has a minimum value of 34.1% in September 2020 and increases to 80.3% by December 2028.