By John T Chapman & Hang Jin, Cisco Systems
The 42 MHz (65 MHz for Europe) upstream return path is running out of bandwidth. The 90 to 100 Mbps it has today is enough to support the 1 Gbps DOCSIS downstream of today, but it is not enough to support the oncoming 10 Gbps DOCSIS downstream of tomorrow.
That means that the entire HFC plan will need to be upgraded to a new return path. There are multiple options available: 85 MHz and 204 MHz using classic FDD and up to 684 MHz using FDX DOCSIS. 85MHz only offers 400 Mbps of bandwidth which could support a 4 Gbps downstream, assuming a 10:1 ratio in bandwidth, but is still not enough for a full 10 Gbps downstream.
FDX DOCSIS offers the most bandwidth possible in the return path with the least impact on forward path bandwidth. However, the cost of deploying FDX DOCSIS today is high as FDX DOCSIS currently is specified to work on an N+0 HFC plant whereas the 204 MHz does not required N+0. However, 204MHz does require upgrading every single amp as well as the node. What if FDX DOCSIS could be deployed by also just upgrading the amps and nodes, rather than trenching new fiber for new node locations? This would make FDX DOCSIS be much closer to cost parity with 204 MHz. In fact, it could save operators billions of dollars. This white paper will show how FDX DOCSIS can be taken beyond N+0. It will do so by describing a deployment model that scalable and allows the cable operator to trade-off cost versus performance.