Optical access networks provide a cost-effective solution to meet the ever increasing bandwidth needs in mobile radio access networks. The main role of these network is to serve calls and transport short messages or packet switched data from radio interfaces to mobile core networks. Increasing user bandwidth needs means higher base station densities. Radio access networks will be able handle more users and higher data volumes more and more small cell base stations added to the network, however installing new small cells, especially their transport network links is challenging. Currently, telecom network providers typically have separate fixed access networks, and mobile transport network infrastructure. In the near future, the cost pressure will drive the operators to have converged optical access networks, due to higher data rate needs, profitability and simplified small cell installation. In my work I have analysed possibilities to achieve convergent networks and made a comparison based on their capital expenditures needs.
Firstly, I studied the requirements to be satisfied in mobile transport networks, mainly imposed by state-of-art LTE networks. I made this study both for backhaul and fronthaul networks and the outcomes will be presented in my thesis.
Secondly, I had to obtain deeper knowledge about may be relevant optical networking technologies and architectures, which can satisfy requirements for LTE capable optical networks. Based on accessing shared media and making multilevel distribution network I show optical network technologies for convergent networks which can be implemented in future.
To make cost estimations I had to plan and implement an optical network planning and dimensioning software module in Matlab. This program can estimate cost impacts for several optical network architectures and store data for later processing in easy-to-use format for humans and computers.
Finally I made a comparison for network solutions and took considerations based on estimated system costs. In this study I made cost comparisons in artificial cost units, not in absolute costs (i.e. real currency), because of sensitive nature of cost data.