Skip to main content
Nov 27 2014

Can you use HFC for small cell backhaul?

Hybrid-fiber coax (HFC)/data over cable system interface specification (DOCSIS®) access networks are widely used to address the needs of residential and enterprise consumers for high speed Internet (HSI), video, and voice services. A hot question these days is: "Can HFC/DOCSIS access networks also support the backhaul needs of LTE small cells?" Since the characteristics of cable networks and the products used can vary significantly across markets, it is not always possible to have a clear yes/no answer. In performing techno-economic analyses for HFC -based backhaul applications, we have seen concern about meeting throughput and performance (packet delay/jitter/loss) requirements. And these can vary with the specific DOCSIS QoS mechanisms and bonding configurations.

SMALL CELL REQUIREMENTS

Small cells are fundamentally used to:

  • Increase wireless network capacity in traffic ”hotspots”
  • Provide coverage in areas where there is poor macro cell coverage
  • Provide both capacity and coverage in targeted high-value areas

Operators are primarily motivated to deploy small cells to deliver higher user experience with better economics. Small cells are best suited for serving applications with high quality of service (QoS) requirements (video streaming, video calling, etc.) which need high throughput and low delay/jitter/packet loss. Environmental Small cells are typically used in indoor as well as outdoor environments. The type of environment has a major correlation with the anticipated user demand, and hence, the potential bandwidth requirement for traffic backhaul. Throughput Backhaul bandwidth requirements will depend upon:

  • Spectrum used by the small cell (typically 5, 10, or 20 megahertz - MHz)
  • Number of sites
  • Use of hub or daisy-chain topology

Performance Many performance requirements -- such as packet delay, jitter and loss for voice, data and video -- do not present a fundamental challenge for an HFC/DOCSIS access network. On the other hand, control signaling performance requirements for LTE radio access networks are more stringent. During a handover (one cell handing off to another), the longer the handover preparations phase, the higher the chance of handover failure. This is a primary area of concern for HFC/DOCSIS-based backhaul.

HFC/DOCSIS NETWORK CHARACTERISTICS

Throughput HFC/DOCSIS 3.0 provides a channelized access network architecture. Downstream channels for data services typically achieve ~38 Mbps (@6 MHHz) or ~50 Mbps (@8 MHz ) per channel. An upstream channel may achieve typically ~28 Mbps (6.4MHz). Today there are only a handful of upstream channels, while there are well over 100 downstream channels. To improve aggregate throughput, channels can be bonded together and a service group concept used to define how DOCSIS channels are allocated to serve consumers over a defined geographic area. Note that dimensioning the upstream on the HFC/DOCSIS access network to support small cell backhaul is complex given the limited number of available upstream channels and lower throughput per channel. Also, should service group splits be required to reduce the size of the geographic serving area, this implies network changes and costly additional HFC/DOCSIS equipment. Performance Our analyses show important factors that contribute to performance issues in HFC/DOCSIS access networks include:

  • DOCSIS QoS – although DOCSIS QoS mechanisms can be used to give preferential treatment to control traffic from small cells, the choice of the QoS mechanism can impact backhaul performance.
  • Use of a single or bonded DOCSIS channels – slight increases in delays may be incurred when channels are “bonded”.
  • Equipment used – design, implementation, and device components used can contribute to equipment performance differences between DOCSIS vendors.
  • Consumer traffic – contention for bandwidth by residential and enterprise consumers, or by multiple small cells themselves.
  • Fiber and coax lengths – while not a large contributor to delays, excessive lengths contribute to propagation delays in the HFC and IP networks.

OUR ANALYSIS

We have seen that the feasibility of using HFC/DOCSIS access networks for traffic backhaul from small cells depends on many factors including:

  • Use case
  • Traffic mix
  • Environment
  • Peak busy hour demand
  • HFC/DOCSIS network characteristics
  • Performance objectives
  • QoS mechanisms

There are currently no commercial large-scale examples of HFC/DOCSIS access networks being used to backhaul traffic from small cells or macro cells. The feasibility and design/cost tradeoff of using HFC/DOCSIS networks for backhaul, however, can be determined through a detailed techno-economic analysis.

Related Material

DOCSIS® for LTE small cell backhaul white paper

To contact the author or request additional information, please send an email to networks.nokia_news@nokia.com.

About Marty Glapa
37 years of experience in Telecommunications, with a primary focus on cable networks, and access networks; leading network planning, modeling, and forward looking projects. Lead corporate interface to CableLabs® for participation in technical forums defining new cable industry architectures and specifications, and inter-operability event participation. Spearheading a next generation distributed intelligence and functionality access architecture/technology evolution and total cost of ownership analysis. Designed, developed and delivered multi-faceted strategic planning war-games that identified and proposed solutions to address network and business challenges, resulting in service provider acceleration of broadband deployments. Analyzed the application of wireless small cells to solve capacity and coverage problems in wireless networks based on operating spectrum. Led analysis investigating deploying a wireless network using macrocells, metrocells, and site sharing partnering. Researching a process and tool-set blending scenario planning with network and business modelling to address innovation challenges and major uncertainties, reduce risk, improve decision making, and stimulate innovation. Oversaw research and prototyping innovation initiatives tying network-based application programming interfaces with end-user device execution environments, resulting in first-ever proof-of-concept application demos of the kind.