Summary
Termination of a telephone call can only be realized by the network operator of the receiving party. This is frequently referred to as "termination monopoly". And accordingly the markets for fixed and mobile call termination are regulated ex-ante including price control. The current regulatory regime considers only those parts of the network where customers compete for jointly used resources (mainly bandwidth) to determine the costs of call termination. Therefore, the critical border is the demarcation point between the dedicated access network and the aggregation network where customers compete for bandwidth.
In PSTN/ISDN networks this demarcation point was in concentrating units usually collocated with the Main Distribution Frames (MDF). However, when applying this regulatory framework to Next Generation Access (NGA) networks the location of the demarcation point could be located at different network levels depending on the NGA architecture considered, leading to heterogeneity of termination cost. In order to discuss these problems on a more solid basis we have set up this study to analyze qualitatively and quantitatively the impact of migrating to NGA on the cost of termination when applying the current methodology for cost determination.
First, we analyzed five NGA-architectures (Fibre-to-the-Cabinet, Fibre-to-the-Building, Fibre-to-the-Home in both Point-to-Point and Passive Optical Network technology as well as Hybrid Fibre Coax (broadband cable networks)) with respect to the location of the demarcation point. This proved heterogeneous developments: In the majority of architectures the demarcation point moves closer to the end user, i.e. the dedicated part of the network is increasingly getting smaller which means that more network elements should be considered for the calculation of call termination costs. However, in the case of Fiber-to-the-Home in Point-to-Point technology the part of the network relevant for termination cost can be shortened when current MDF locations are not used for concentration anymore. It proved to be especially difficult to identify a uniform Demarcation Point for Passive Optical Networks (PON), among other issues because of the large variety of splitter deployment alternatives.
We then expanded our WIK NGA cost model in order to approximate the termination cost for Fiber-to-the-Curb and Fiber-to-the-Home networks of a hypothetical German network operator. In the base scenario with voice and data usage on a current level it could be shown that termination cost was higher when additional network elements of the former local loop segment had to be considered in the calculation due to the location of the demarcation point being closer to the customer. However, sensitivity analysis also showed that termination cost of all architecture variants increasingly approaches the level of FTTH/Point-to-Point when traffic volume and share of non-voice services increases. When IPTV is included, this level differences between the architectures because of IPTV’s relatively high traffic share in those network segments with high investments. We discuss alternatives when applying the methodology and evaluate results critically.
(Full version only available in German language)
Discussion Paper is available for download.
