By Jeroen Wellen, Prudence Kapauan & Amit Mukhopadhyay, Bell Labs Consulting/Nokia
The access network is the costliest investment for any operator and it is also the hardest to evolve to next generation technologies. Incumbent operators try to prolong their investment with minimal alterations to the “last mile” connection while new operators try to make their investments future proof as far into the time horizon as practical. While Fiber to the Home (FTTH) is the unquestioned leader from a performance perspective, the economics of the solution works out only in select morphologies and market conditions. Access technologies like Hybrid Fiber Coax (HFC) and Digital Subscriber Line (DSL) comprise the majority of the residential broadband connections today and Fixed Wireless Access (FWA) is also gaining momentum in certain markets. The challenge for DSL is that it can match FTTH performance for only limited loop lengths. HFC is largely constrained by the fact that it operates in a shared medium environment. FWA shares both the limitations – distance as well as shared medium.
The key is to balance the market-specific technology needs and the business justification of the capital and operational expenses. In other words, a technology that may be suitable for delivering the services to a particular geographic and demographic morphology, may be completely unsuitable for delivering a similar service to another morphology. Access networks are considered long term investments and operators make decisions based on revenue potentials and thus Return on Investment (RoI) over a relatively longer period of time is generally acceptable.
Choosing the right Access Network technology at the right moment is one of the most challenging decisions faced by Network Operators. How long the installed base can support the ever-growing demand of applications, especially video, and what technology to install or upgrade to, and when, are the key questions to be answered. HFC and FWA are particularly attractive technologies since they connect multiple homes to a single Access Node (a fiber node in HFC and a cell site for FWA) through a shared medium, allowing to benefit from stochastic multiplexing while minimizing drop costs. But the advantages of these technologies depend on how many homes can be connected per node. And, to assess the techno-economics of the solution, proper estimation of the expected traffic is needed – at present and during the foreseen lifespan of the investment.
This paper presents a model to estimate the 10 years maximum sustained throughput requirements for residential broadband services. The model is used to estimate the maximum through puts during busy hour and to assess how upper throughput percentiles compare to the mean expected value. In addition, the-impact of other effects such as disruptive video applications and loads during special events are investigated. The main purpose of the paper is to provide guidance to operators investing in either upgrading their current network or investing in new network technology. While multi-gigabit access connections often catch the headlines, that is not the sole criteria around which access networks need to be designed. It is imperative that shared medium technologies provide such peak speed connections to support demanding applications. In many circumstances, it is equally important to provide high quality sustained throughput connections at considerably lower speeds. The service targets and the choice of technologies are driven by market economics.