Why DFB Lasers May Be Insufficient in DOCSIS 3.0 Upstreams

Cable operators that want to maximize the value of their hybrid fiber coax (HFC) networks are deploying DOCSIS 3.0 channel-bonding technologies to increase capacity and to compete with telcos’ all-fiber networks. However, there are two drawbacks when it comes to using DOCSIS 3.0: Upload speeds usually are much slower than download speeds, and noise is a problem in the upstream.

There are many options for improving the carrier-to-noise ratio (CNR) in the upstream in DOCSIS 3.0 systems. Aurora Networks recommends digital return technology and makes its case for this in a white paper "DOCSIS 3.0 Solving the Upstream Challenge."

According to the white paper, "For DOCSIS 2.0 implementations, driven by the need for higher-speed data services to compete with the telcos and their long upstream cascades, cable operators were forced to segment their networks, driving down the serving area size and hence the load on the upstream link. For DOCSIS 3.0, operators are being forced to combine between four and eight upstream optical links to ensure that they are optimizing their CMTS investment."

Optical-link combining leads to degradation in CNR performance, the paper says, adding when longer distances need to be covered, analog links must be cascaded, which leads to further CNR degradation.

In the past, operators deployed analog Fabry-Perot lasers and, later, analog distributed feedback (DFB) lasers for transmitting. But in DOCSIS 3.0 systems, with four or more bonded upstream channels, these are not sufficient, stated John Dahlquist, vice president/Marketing with Aurora Networks.

"I do agree that Fabry-Perot is probably not a good idea," said Ron Hranac, technical leader/Cable Access Business Unit at Cisco and senior technology editor for Communications Technology magazine. "More than one or two channels in the upstream is pretty much the limit of Fabry-Perot. With DOCSIS 3.0, they’ve got to do something with those old Fabry-Perot links."

Hranac said most operators are replacing Fabry-Perot with analog DFB lasers or digital returns, but Aurora’s paper posits that "analog DFB transmitters can support DOCSIS 3.0 under limited – reduced reach, reduced combining – scenarios, but will not prove suitable for future generations of DOCSIS with higher-modulation schemes and for higher loading."

"Digital return technology is the optimum solution," noted Dahlquist.

The white paper backs this up, saying, "Digital return technology provides full support for DOCSIS 3.0 under all conditions, including multiple levels of combining and increased channel loading, and will support higher modulations schemes (256-QAM) to support future generations of DOCSIS. This is the future-proof solution."

-CT Staff