Advertisement

Ever since 1987, players from the winning Super Bowl team have been asked the question, “You've just won the Super Bowl, what are you going to do next?”  They respond, “I’m going to Disney World!”  This led me to wonder how the U.S. Treasury would answer the same question after the conclusion of the latest AWS-3 auction in which Mobile Network Operators bid almost $45B to secure new licenses – a number that easily surpassed analysts’ estimates of between $10B and $20B.  Why? As highlighted by the December 2014 edition of the Ericsson Mobility Report, mobile subscribers are expected to drive a nearly 8x increase in mobile data (particularly mobile video) between 2014 and 2020. As a result, mobile networks need more capacity, and the way to get more capacity is to 1) deploy more spectrum, perhaps in conjunction with carrier aggregation and 2) densify with more cell sites, especially small cells.

The question for mobile network operators now becomes, “How do you deploy your costly new spectrum in a way that ensures you have the highest capacity and lowest cost network possible.” After spending a record $45B on spectrum, it is now time for mobile operators in North America to give serious consideration to the C-RAN architecture, and in particular, a C-RAN architecture enabled by an OTN-based fronthaul network.

Deploying capacity in the traditional distributed manner, where the baseband for each sector is located at the cell site has serious disadvantages in the LTE-A era and is most certainly counter to the requirements of 5G. The disadvantages include:

  1. Inefficient use of the baseband. On average, utilization is only around 30 percent because they need to have enough processing capability to handle the peak load for the area served.
     
  2. Limited, if any, ability to take advantage of new LTE-A features like Coordinated Multipoint for interference reduction.
     
  3. High power consumption, up to 46% of which is for non-radio-related items like air conditioning.
     
  4. Challenging site dynamics and increased costs, including where to put the new radios and baseband, site preparation and on-going lease expenses.

…and the list goes on.

On the other hand, the C-RAN architecture not only alleviates each of these considerations, but it enables operators to take the first and perhaps most important step toward a cloud-enabled mobile network. The fundamental concepts behind C-RAN are 1) move the baseband processing resources for all cell sites and small cells within an area to a centralized location, and 2) (eventually) leverage NFV/SDN to virtualize the baseband processing onto cloud infrastructure. 

Click to open full view

Figure 1 – Advantages of the C-RAN model

C-RAN has several important technical advantages. For example, by co-locating baseband units, support for interference busting techniques like Coordinated Multipoint (CoMP) can be employed. In the paper, “Recent Progress on C-RAN Centralization and Cloudification,” China Mobile has shown uplink CoMP gains of up to 100 percent at the cell edge.  As a mobile operator, you may have just spent billions to acquire new spectrum. Doesn’t it make sense to make the most efficient use of that spectrum throughout your footprint, including at the cell edge (which, as the network is densified, there will be more of)?  

C-RAN Drives the Need for a Mobile Fronthaul Network and OTN is the Enabler

The fundamental challenge to unlocking the advantages of C-RAN is the rollout of a cost-effective, fiber-efficient, and scalable “Fronthaul” network that interconnects the radios at the cell sites with the centralized baseband units. Mobile network operators have a few options for how to do this.

Passive and Active WDM solutions are available, but they lack carrier-grade OAM, fault isolation and demarcation that should be a requirement in every mobile network operator’s Radio Access Network. They also require an overinvestment in higher cost WDM transceivers.

Interest is growing around how to use Ethernet for Fronthaul. While Ethernet is attractive from the sense that it enables the use of carrier Ethernet infrastructure, we’re likely several years away from standards and technology being in place to enable its widespread adoption in the RAN.

This leaves the ITU-T-standardized G.709 OTN (Optical Transport Network). The following table highlights the advantages and challenges for each of the fronthaul methods.

Click to open full view

As the table highlights, OTN based fronthaul offers many advantages to the mobile operator:

  1. It is the most fiber-efficient option, offering 4x the fiber efficiency of passive or active WDM.

     

  2. It features carrier-grade and standardized OAM.

     

  3. OTN-based equipment provides a natural demarcation point between the wireless operator and the wireline/fronthaul network operator.

     

  4. The combination of standardized OAM and demarcation enables important OpEx reducing characteristics like simplified fault isolation.

     

  5. It provides an opportunity to enable an SDN-controlled fronthaul network that:

     

    1. Gathers statistics about each connection and each link, including BER, as well as uplink and downlink delays;

       

    2. Remotely configures each node and into which OTN container each CPRI link should be mapped; and

       

    3. Enables load-based sizing of the fronthaul traffic. CPRI links from small cells or secondary carrier/sectors can be selectively enabled or disabled for fronthaul transport.

       

  6. An OTN-based fronthaul network can be shared by multiple operators, and it can transport multiple protocols including FE/GE simultaneously.

     

  7. OTN as a technology is well understood by carrier operations teams.

     

The combination of these advantages means that OTN represents the most cost-effective, fiber-efficient, and scalable “Fronthaul” network alternative.  The challenge, has been a lack of OTN-based fronthaul solutions which meet the strict 3GPP and CPRI latency and timing requirements. However, PMC recently announced an OTN fronthaul solution, based on their HyPHY 20Gflex OTN Processor, which enables low-latency multiplexing of CPRI Option 1-5 signals into 10G OTN-based optical links.  OTN fronthaul of CPRI clients with the HyPHY 20Gflex beats the frequency accuracy requirement of the CPRI standard by 75%, has zero impact on EVM performance, and achieves sub 3ns latency variation.   Carriers can now realize the benefits of OTN-based fronthaul without sacrificing network performance. 

 

An operators’ Radio Access Network is arguably its most important asset. It is the superhighway that it uses to connect to its subscribers.  With the availability of carrier-grade, cost-efficient, and scalable OTN based fronthaul technology, why trust your most important asset to anything else?
 

 

 

Scott Wakelin is a Senior Product Marketing Manager in the Communications Business Unit at PMC-Sierra and is responsible for OTN and mobile fronthaul solutions and strategy. He has more than 15 years of experience delivering innovative semiconductor solutions to the networking industry. Recently, Wakelin was recognized with PMC’s 2014 “Winner’s Circle” Award for outstanding achievement. He holds a Master of Applied Science degree and is a member of the Marketing Strategy “Special Interest Group” within the Center for Sales and Marketing Strategy at the University of Washington’s Foster School of Business.

Advertisement
Advertisement