Bullpen: The Big Gun Idea

The cable industry’s narrowcast-generated services coupled with strategic implementation of time and frequency division multiplexing (TDM and FDM) has created an unrivaled and sustainable network.

Yet this current converged network of voice, video and data is an incoherent mix of TDM and FDM with layers of hierarchical combining, comprising excessive numbers of wires, which creates an infrastructure that is hard to sustain and limits the velocity of service evolution. The dichotomy exists where the most valuable services are being deployed in a way that could prove to be the hardest to manage.

Convergence does exist in the GigE networks of dense wavelength division multiplexing (DWDM), routers and switches; but it has yet to be translated into the narrowcast.

The Big Gun concept fully embraces the digital revolution — where processing speeds already far outstrip real-time video — and promotes the idea of targeting the end game instead of waiting on Moore’s Law. Sometimes it is better to set a firm direction than to just ride the technology curve.

"Wiring and re-wiring narrowcast services…is analogous to painting the Golden Gate Bridge."

TDM-to-FDM transform

Today cable is "all-digital IPTV" where video signals are transmitted in the TDM domain. It is transformed to FDM domain for delivery within the HFC plant. Either discrete or dense devices generally known as edge-QAM modulators perform the transformations.

The transformations occur in single channel or block up-conversion of multiple adjacent channels in varying and ever increasing blocks (2, 4, 8, 16, or 32). The most advanced technologies perform comb-channel insertion where single channels or blocks of channels are inserted anywhere within the spectrum.

While these technologies are all advancing, I believe that they are solving "the hard problem" instead of focusing on the "right" problem.

Simply put, the Big Gun is a single TDM to FDM Fourier transform device that does the following:

• Generates the entire spectrum (54 MHz – 1GHz)

• Outputs RF or direct laser out

• Provides a single RF Gun per service group providing all services

• Supports VOD and SDV protocols and shares bandwidth across services

• Supports a DOCSIS Timing Interface (DTI) for high-speed data (HSD)

• Relies on external "bulk encryptors" for broadcast, common-tier and session-based encryption

This approach does not mean that hierarchical service groups disappear or are not supported; rather they are virtualized in the TDM domain and subscribed to by the edge device that performs the transformation to FDM domain.

The historical FDM combining is done in the virtual TDM domain by the use of advanced digital signal processing (DSP), field programmable gate array (FPGA) and digital to analog converter (DAC) technologies instead of conventional RF combining with physical wires. This allows future "re-wiring" to be an exercise of software re-configuration rather than late-night trips to the hub sites.

Consider the range of change — all affecting the narrowcast — that lies ahead for the industry: continued analog reclamation, growth of on-demand services and libraries, launch of StartOver or other real time acquisition (RTA) services, remote storage (RS) DVR, switched digital video (SDV) service group sizing, sizing of tuners and QAM signals per SDV service group, placement of content in the switched tier, growth of HSD and voice, continued launch of DOCSIS 3.0, DOCSIS Set-top Gateway (DSG) standardization, launch of IPTV, etc.

Admittedly, the desired end-solution of natively supporting every service may not be practical because of complexities of different modulation schemes and timing, specifically analog and RF-two-way.

But the use of front-end combining, internally creating the digital, VOD, SDV and CMTS spectrums and then generating a full RF spectrum or one in a direct laser output to the node can still realize tremendous gains.

No more combining?

Hierarchal relationships (parent/child/sibling) across multiple services are hard to manage. Cascading combining networks consume tremendous amounts of RF power, requiring QAM modulators to have higher per-channel RF output and thus driving up space, electrical, and cooling requirements, and in the extreme case, necessitating new hubsites.

All the while, there remains the chore of wiring and re-wiring narrowcast services, which is analogous to painting the Golden Gate Bridge: labor intensive, complicated and continuous.

The final RF spectrum generated by the Big Gun could theoretically be nearly perfect. Eliminated from the generation path would be many of the steps of generating the spectrum that impact signal quality. The basic RF level out of this device (not using direct laser out) would only have to be about +17 dBmV to go directly into the laser.

When the convergence of the network occurs at the transformation point of TDM to FDM the complexity of the physical plant will diminish through the process of simplification and reduction.

Space, the final frontier

Apart from network complexity, the lack of space, electrical and cooling that can constrain growth or impede the rollout of additional services argues for a TDM-to-FDM transformation.

The Big Guns will consume less current and space and require less cooling than the existing narrowcast edge devices, while providing a full spectrum inclusive of all services. Estimates are that a Big Gun might consume about 0.5 ampere and require ½ a rack unit (RU = 1.75 in) to generate a complete spectrum per service group.

Note that a typical rack has 42 RUs and consumes roughly 1 ampere per RU, or 2 amperes if there is redundancy. A hubsite supporting 80 service groups may need 20 full racks to generate the spectrums used today. With the Big Gun, that might fit within a single rack or two. Of course, lower electrical consumption means less cooling.

Then there are the literally thousands upon thousands of wires, specifically coaxial cables, within a hubsite. The disparate hierarchies between services of service groups create complexities that evade — consider late night wiring sessions — the best management intentions of spreadsheets and cable labeling.

There are just too many wires out there. Their time has passed. The Big Gun simplifies and reduces the problem and the wires drop out of the solution.

Fire — and recoil?

To sum up, the Big Gun:

• Resolves hubsite environmental constraints of space, electrical and cooling

• Simplifies and reduces narrowcast complexity

• Effectively ends narrowcast re-wiring

• Provides a simple redundancy & resiliency N+1 architecture for services

• Provides a configurable platform supporting accelerated service velocity

• Future-proofs the platform through FPGAs and DSP

The Big Gun approach will optimize the combination of TDM and FDM. In terms of access control, authorization, and encryption, it can take advantage of all the capabilities of TDM and maximize them via FDM. This unified architecture will ensure a competitive platform into the future.

While not discussed here, the return path is equally important. Next up: "Recoil," the Big Return Gun, a unified return path device….

Glen Hardin is chief architect, video systems, for Time Warner Cable.