as a sales engineer its important to keep track of where the service provider industry is heading over the short-term and long-term from a technology perspective. by staying engaged you can better define where training and research need to be dedicated in order to be successful at engineering solutions. basically, allowing old stuff to be purged and knew stuff to be stored.
in the current climate of broadband access service providers have 4 options, with various sub-options, but 4 main options for deploying broadband services to their customer base: docsis 3.0, 4G LTE, VDSL2, and fttx. all of these technologies must use fiber, leveraged in strategic and cost effective manners, to be successful. docsis uses fiber to feed nodes out in the field. these nodes have coaxial connections out to homes. by limiting the number of homes passed per node, a cableco can increase overall bandwdith to those homes. in a 4G LTE environment, fiber is used to connect each cell tower back to a CO or headend for data routing, voice termination, or wholesale backhaul. in order to provide competitive speeds with VDSL2, remote cabinets must be placed close to the residence, know as fiber to the neighborhood or curb. From these cabinets copper is connected to each home. with vdsl these cabinets must be as close to the customer as possible in order to provide competitive speeds as compared to docsis and fttx. with fttx, I think the the implications of fiber are obvious – of course the amount of fiber is dictated by whether a gpon or active ethernet architecture is being deployed.
dsl prime published a short but interesting article on the theoretical and not so theoretical reality of docsis 3.0 bonded speed potentials in their docsis reports section. by bonding a larger and larger subset of RF channels the amount of shared bandwidth potential is high. cable providers that focus on pushing fiber further out in their network, performing node splits to decrease the number of homes competiting for bandwidth, and removing analog channels to open RF channels for bonding are best positioned to capitalize the bandwidth race.
engineers involved in the field should be focused on cmts solutions such as cisco’s ubr10k, rf gateway, and 3g60 line cards. from a headend perspective i would focus on solutions and equipment that allow me to remove all traces of analog to free up rf space for bonded channels and eventually to a vdoc solution, and combining changes for node splits. in the outside plant i would focus on splitting nodes and upgrading plant capacity to 1GHz to give the maximum potential bandwidth on the plant.