Dynamic LTE network dimensioning -

Abstract

Radio network dimensioning is an essential step towards designing a cellular network and has traditionally relied on a pragmatic approach to capture the effects of fading and shadowing on the system interference. However, with the increased flexibility of LTE, which builds on the statistical variation of wireless channels by adapting resources accordingly, there is a need for new customized dynamic dimensioning methods. In this thesis, we develop the Statistical Link Budget Analysis (SLBA) approach for LTE dimensioning that jointly accounts for fading and shadowing in the channel as well as loading due to traffic demand in the interference calculation. We utilize the derivation of the probability density functions (PDF) of four dimensioning parameters: SINR, noise rise, maximum allowed path loss and cell range in closed form. Results are generated and analysed to capture the LTE trade-off between coverage, capacity, and quality constraints and the findings show tight dependency which invites a form of iterative process to optimise performance. Accordingly, an iterative algorithm that balances coverage and capacity requirements while respecting quality target is developed and implemented in a VBA-Matlab tool: Statistical Analysis of LTE Dimensioning (SALTED). The algorithm is tested in a case study and results are benchmarked against traditional LTE dimensioning approaches with the help of a professional LTE network planning tool. The outcomes form this exercise show that the traditional approach leads to over-dimensioning at extra cost, thus, advocate the advantages of SALTED which yields dimensioning results comparable to those obtained with the professional planning tool.