Sort by:[Date]

IHS-Infonetics: Optical Network Equipment Spending Trends + Huge Growth for 100G Optical Ports

IHS-Infonetics just released vendor market share and preliminary analysis from its 2nd quarter 2015 (2Q15) IHS Infonetics Optical Network Hardware report (Full report published by August 24th).   According to the report, global optical network hardware revenue (WDM and SONET/SDH) grew 22% sequentially in 2Q15, but was flat on a year-over-year basis. Europe is the only major world region that posted positive year-over-year growth in 2Q15, up 8%.  


  • On a rolling 4-quarter basis, WDM equipment spending further extended 3 years of consecutive growth
  • Spending on WDM equipment grew 23 % in 2Q15 from 1Q15, and was up 6 % from 2Q14
  • WDM gear comprised 86 % of total worldwide optical hardware revenue in Q2
  • Spending on optical network hardware in Asia Pacific surged 36% in 2Q15 from the previous quarter, but is down 2% from one year ago
  • Alcatel-Lucent announced an intention to merge with Nokia, an action IHS does not expect to have any transformative effects on ALU’s optical business or the competitive landscape



Analyst Quote:

“With three consecutive quarters of good results under its belt, Europe is signaling a reversal of the terrible optical spending that we’ve seen in the region over the last five years,” said Andrew Schmitt, research director for carrier transport networking at IHS. “This strength is concentrated in Alcatel-Lucent, Ciena and Infinera.” 

"When taking into account currency effects, the results are even stronger - adjusted for exchange rate, optical spending in Europe saw a 30 % year-over-year growth rate in the second quarter when measured in euros," Schmitt said.


The quarterly IHS Infonetics Optical Network Hardware market size, share and forecast report, led by analystAndrew Schmitt, examines the vendors, markets and trends related to metro and long haul WDM and SONET/SDH equipment used to build optical networks. The report also tracks Ethernet optical, SONET/SDH/POS and WDM ports. Vendors tracked include Adtran, Adva, Alcatel-Lucent, Ciena, Cisco, Coriant, Cyan, ECI, Fujitsu, Huawei, Infinera, NEC, Padtec, Transmode, TE Connectivity, Tyco Telecom, ZTE, others.

To purchase the report, please visit




Separately, IHS-Infonetics reports that Coherent 100G port shipments for metro regional optical networks grew 145 % in 2014 from the prior year, and are anticipated to grow another 118 % in 2015.

“Adoption of 100G coherent technology has surged, first in long haul networks and now becoming a material part of metro networks. The expansion of 100G into new markets was the catalyst for our 100G+ coherent optical ports report, which provides an accurate, in-depth picture of how 100G technology is being used today and how it’ll be used in the future as the landscape grows increasingly complex,” said Andrew Schmitt, research director for carrier transport networking at IHS.

“100G is poised to explode in 2016 as new equipment built specifically for the metro reaches the market, allowing 100G technology to economically reach new portions of the network such as metro edge and metro regional,”Schmitt said.


  • 2014 was a banner year for 100G port shipments, led by massive purchases in China from China Mobile
  • Most of 100G coherent technology deployed in 2014 was for long haul applications, but metro regional (<600km) and metro access (<80km) applications will start ramping in 2016
  • 100G market share is concentrated in a small circle of players: Alcatel-Lucent, Ciena, Huawei, Infineraand ZTE; the only potential catalyst for shifts will come from deployment in shorter reach metro and datacenter applications—the next growth vector for 100G
  • Sometime in 2017–2018, 100G coherent will make another quantum jump, displacing 10G in the 80km or less metro-access market




The 20-page IHS Infonetics 100G+ Coherent Optical Equipment Ports market size, vendor market share and forecast report provides detailed granularity for 100G+ coherent and non-coherent port shipments on optical transport equipment, tracking the evolution of 100G as operators increase the flexibility and capacity of their networks. The report tracks 100G by application, including metro regional, metro access and long haul, as well as specific technology derivatives such as flex-coherent and direct-detect 100G.

To purchase the report, please


Competing market research firm Dell'Oro Group says the optical transport network equipment market will grow at a 10% compounded annual growth rate (CAGR). Demand for metro WDM capacity will drive up the overall optical transport revenues to $15 billion by 2019 according to this Dell'Oro report.

Exposed url:

During this four-year period, service providers will continue to deploy a mix of 100G and 200G wavelengths in their networks. The research firm forecast that over 75 percent of WDM capacity will be from 100G wavelengths, while 200G will contribute nearly 25 percent of WDM metro equipment revenue by 2019.

Jimmy Yu, VP of optical transport market research at Dell'Oro Group, said in a release that the majority of metro equipment sales will come from traditional service providers, but content providers and financial trading companies will contribute to overall growth as they install their own 100G networks.

"The majority of metro equipment purchases will still be made by telecom service providers, expanding their metro network capacity for higher speed services, but we also see a strong trend towards enterprises such as Internet content providers and financial institutions procuring and installing their own high speed 100 Gbps links. This trend is being powered by the increasing importance of data centers to a company's core business," Yu said.

"The network still needs a lot of raw bandwidth and WDM is the best equipment to deliver that.  While high demand for long haul equipment will continue, the biggest growth that we are projecting is in metro applications. The majority of metro equipment purchases will still be made by telecom service providers, expanding their metro network capacity for higher speed services, but we also see a strong trend towards enterprises such as Internet content providers and financial institutions procuring and installing their own high speed 100 Gbps links. This trend is being powered by the increasing importance of data centers to a company's core business," added Mr. Yu.


Verizon conducts field trial of 10 Gb/sec Nex Gen PON2 service; ITU-T rec's for NG PON2

Verizon has completed a field trial of NG-PON2 fiber-to-the-premises technology that could provide the infrastructure for download speeds up to 10 Gbps for residential and business customers. The huge telco's current top download speed for its residential FiOS service is 500 Mbps. The new technology could "open the door" to speeds as high as 80 Gbps, according to Verizon. 

The field trial took place on a network link between the company's central office in Framingham, MA and a home three miles away served by Verizon FiOS. The test required installation of a new optical line terminal (OLT) at the central office supporting four wavelengths, each capable of delivering speeds up to 10 Gbits/s downstream and 2.5 Gbits/s upstream. Verizon also said it was able to demonstrate the simultaneous use of standard GPON and NG-PON2 on a single fiber, and a successful fail-over scenario where its new ONT autonomously restored 10G service by tuning to a new wavelength after a simulated fault was introduced.

Vendor partners in the trial included Cisco Systems Inc. and PT Inovação (part of the Portugal Telecom Group) which provided the NG-PON2 equipment system.

For Verizon's upcoming NG-PON2 RFP, there are vendors like Alcatel-Lucent, Adtran Inc. Calix Networks Inc Huawei Technologies Co. Ltd. and of course Cisco. 

"The advantage of our FiOS network," said Lee Hicks, vice president of network technology at Verizon, "is that it can be upgraded easily by adding electronics onto the fiber network that is already in place. Deploying this exciting new technology sets a new standard for the broadband industry and further validates our strategic choice of fiber-to-the-premises."


ITU-T recommendations for NG PON2 specify up to 40G b/sec speed:

Recommendation ITU-T G.989.1 series describes 40 Gigabit-capable passive optical network (NG-PON2) systems to an optical access network for residential, business, mobile backhaul, and other applications.

Recommendation ITU-T G.ngpon2.1 addresses the general requirements of 40 Gigabit-capable passive optical network (NG-PON2) systems, in order to guide and motivate the physical layer and the transmission convergence layer specifications. This Recommendation includes principal deployment configurations, migration scenarios from legacy PON systems, and system requirements that are requested by network operators. This Recommendation also includes the service and operational requirements to provide a robust and flexible optical access network supporting all access applications.

The physical layer specifications for the NG-PON2 physical media dependent (PMD) layer is described in Recommendation ITU-T G.989.2 (ex G.ngpon2.2, draft). The transmission convergence (TC) layer is described in ITU-T Rec. G.987.3, with unique modifications for NG-PON2 captured in Recommendation ITU-T G.989.3 (ex G.ngpon2.3, draft). The ONU management and control interface (OMCI) specifications are described in ITU-T Rec. G.988 for NG-PON2 extensions.






IHS-Infonetics Survey: Network Operators Reveal SDN Plans,Timing & Challenges/Alan's Take

IHS-Infonetcs latest report:  "SDN Strategies: Global Service Provider Survey" of worldwide carriers says that global network operators are moving toward software-defined networking (SDN). The carriers surveyed represent 49% of the world's telecom capex and 46% of global telecom revenue.  

-->The study found that 82 % of service provider respondents have either already deployed SDN, are now deploying SDN, or plan to evaluate it in 2015.


  • The #1 reason service providers are investing in SDN is to simplify and automate service provisioning, which they believe will lead to service agility and quick time to revenue.
  • Various barriers are becoming more prominent as operators get closer to commercial deployment; respondents to this year's survey cited integrating SDN into existing networks and immature technologies and products as the top 2 barriers.
  • Operators want SDN in most parts of their networks, with the top domains for deployment consisting of cloud services offered to customers, within and between data centers, and access for businesses.


Lead Analyst Quotes:

"The successful field trials and a few commercial deployments of SDN in the last year keep moving toward more commercial deployments in 2015, still mostly on a limited basis as operators put one or two use cases to the test under real-world conditions in their live networks," said Michael Howard, senior research director for carrier networks at IHS & co-founder of Infonetics.  

"Carriers are starting small with their SDN deployments and focusing on only parts of their network to ensure they can get the technology to work as intended. We see in the results of our SDN survey that though momentum is strong, it will be many years before we see bigger parts or a whole network that is controlled by SDN," Howard said. 



  • The 25-page 2015 IHS Infonetics SDN Strategies: Global Service Provider Survey is based on interviews with purchase-decision makers at 28 incumbent, competitive and mobile service providers from EMEA, Asia Pacific and North America that have evaluated or deployed SDNs in their networks or plan to do so. IHS asked operators about their strategies and timing for SDN, including deployment drivers and barriers, target domains, use cases and more. 


To purchase the report, please visit

Author's Rebuttal:

We firmly believe that the overwhelming majority of network operators, with the exception of NTT, will not be deploying classical/pure SDN-OpenFlow as standardized by the Open Network Foundation (ONF).  Many are evaluating network virtualization (an overlay model where a logical network is mapped onto a physical network) as per the VMware schema.

However, the vast majority of carriers (and cloud service providers like Amazon, Google, and Microsoft) have invented their own version of SDN and instructed their equipment suppliers to implement that.  In some cases, they use a specific vendor product with some user programmablity, e.g. Cisco Metro Ethernet switches used for AT&Ts Network on Demand service (a switched Ethernet WAN service).

The bottom line is that such carrier or network equipment vendor specific solutions are generally not inter-operable with any other SDN carrier or cloud service provider offering. Hence volumes will be limited and network equipment vendors will need different software for different carriers.  As a result, we will likely see pockets of SDN in carrier WANs, but no mass deployment of ONF standardized SDN-OpenFlow anytime soon!


NFV Market to Grow More than 5-Fold through 2019:

Virtualizing Network Security with NFV and SDN - Whitepaper and Webinar:

35 Percent of Operators Surveyed Will Deploy NFV This Year:

Virtual Routers on Track to Grow 125 Percent in the Next Year:

Data Center and Enterprise SDN Market to Grow More than 15-fold by 2019:

Network Operators Rate Router and Switch Vendors; Cisco #1 for 3rd Straight Year:

Mixed-Bag Carrier Ethernet Equipment Market Set to Top $29 Billion in 2019:

Download the IHS Infonetics 2015 service brochure or log in:
  -  Analyst Note: ONUG Spring 2015: Preparing for Open Networking (June)
  -  NFV Hardware, Software, and Services Forecast (July)
  -  Carrier SDN Hardware, Software, and Services Forecast (Aug.)
  -  Routing, NFV, and Packet-Optical Strategies: Service Provider Survey (Aug.)
  -  Data Center SDN Strategies: Global Service Provider Survey (Aug.)
  -  NFV Vendor Leadership Analysis (2015)

  -  SDN & NFV: Lessons Learned (Sept. 24: Learn more)
  -  SDN & NFV: Accelerating PoCs to Live Commercial Deployment (Watch now)
  -  White Box Switching: Is It Time to Jump In? (Watch now)
  -  Service Provider Experiences with NFV: The Good, the Bad & the Ugly (Watch now)
  -  Evolving Network Architectures: Cloud, SDN, NFV & Packet-Optical (Watch now)
  -  Router Bypass: Using NFV to Deliver Enterprise Services (Sponsor)

IHS Sales: +1 844-301-7334 

New Mobile Virtual Network Operator Service Agreement & Other Wireless Telco News

Wireless communications service providers are licensing their network infrastructure to mobile virtual network operators.   Telco Cuba, Inc. (OTC: QBAN), a U.S. based mobile telecom and data connectivity service provider, announced today that it has immediately begun offering mobile voice and data services to consumers and corporations as a result of a Mobile Virtual Network Operator (MVNO) agreement with Next Mobility.

TelcoCuba now provides high-quality voice and data services to consumers and enterprises utilizing LTE, 4G, and 3G networks. Telco Cuba mobile services will be available in other countries via roaming agreements and services with other mobile service providers.  Telco Cuba and its wholly owned subsidiary Amgentech, Inc. are already well established in the local market for Voice over IP (VOIP) services and communications technology. Amgentech, Inc. has provided services to multiple network operators, constructed its own highly reliable network to enable communications worldwide, and has built highly reliant networks for its client base. Going forward, it will now offer extra-flexible solutions that combine mobile voice and data communication services with existing services for blended communication solutions that include mobile voice and data, mobile VoIP, VoIP, International Dialing, top-off prepaid phone service, international roaming, and much more.

The MVNO agreement signed with Next Mobility is a major first milestone for Telco Cuba. It affords Telco Cuba the ability to enter the cell phone service provider market with a drastically reduced startup cost, allowing Telco Cuba to use its budget where it matters - customer acquisition and marketing. Speed to market is the single most important factor in the digital age. Next Mobility is a well-established entity in the space and our contracted services will afford Telco Cuba a time to market of just under 60 days.

--->Apple is reportedly in talks with telecom companies in the U.S. and Europe to let customers pay the Cupertino-based tech giant for wireless service directly, rather than going through wireless firms like AT&T or Verizon.  The company is conducting private trials of the service in the U.S. and has engaged in discussions with European companies to offer a similar service there, Business Insider reports. Also see related RCR Wireless article.


MVNO – Mobile Virtual Network Operator is a term coined to describe a company that setups a platform for the resell of mobile phone services from one of the big three cell phone providers in the United States of America or elsewhere in the world. AT&T, T-Mobile & Sprint are the biggest of the Mobile Virtual Network Enablers. 


In other wireless and telecommunications news and developments: 

  • Verizon recently announced Grid Wide Utility Solutions, a new Internet of Things (IoT) platform service offering utility companies an easy on-ramp to grid modernization. Now available in the U.S., Grid Wide offers electric utility companies an integrated solution for smart metering, demand response, meter data management and distribution monitoring and control.  With 147 million electric meters in the U.S. today, Verizon's Grid Wide aims to transform the delivery and consumption of energy nationwide for investor-owned, cooperative and municipal utilities and their customers. Designed to maximize the benefits of smart meters, the solution comes equipped with a wide range of cloud-based applications intended to help utility companies drive incremental revenue, reduce operating costs, increase efficiency and improve customer experience.


  • AT&T is promoting a $200 monthly cellphone and TV Everywhere bundle that it plans to offer throughout the U.S. -- marking the first fruits of its merger with DIRECTV. Starting Aug. 10, the carrier will provide four phone lines with unlimited voice and texting, along with 10 gigabytes of sharable data. On the video side, AT&T will hook up four TVs with HD and DVR features and the ability to watch on any mobile device. The New York Times (free-article access for SmartBrief readers) (8/3), CNET (8/2)


  • Vonage Holdings Corp, a leading provider of cloud communications services for consumers and businesses, today announced results for the second quarter ended June 30, 2015.  Second Quarter Consolidated Financial Results - "We continue to drive market-leading growth at Vonage Business, while increasing profitability in Consumer Services," said Alan Masarek, Chief Executive Officer of Vonage.  "At Vonage Business, we delivered 118% revenue growth fueled by the successful execution of our acquisition strategy coupled with strong organic growth. We also made significant investments in our sales infrastructure, brand and leadership team to enhance our position in the rapidly growing Unified Communications-as-a-Service (UCaaS) market." 


  • T-Mobile US Inc last week reported second quarter 2015 results reflecting continued strong momentum, industry-leading growth, and continued low churn. The Company again outperformed the competition in both customer and financial growth metrics. T-Mobile generated 2.1 million total net customer additions, marking the ninth consecutive quarter that T-Mobile has delivered over one million total net customer additions. Additionally, the Company delivered 14% total revenue growth and 25% growth in adjusted EBITDA compared to the second quarter of 2014. 

 "While the carriers continue to use gimmicks to confuse consumers, T-Mobile continues to listen to customers and respond with moves that blow them away," said John Legere, President and CEO of T-Mobile. "On top of adding 2.1 million new customers in the second quarter, we delivered 14% year-over-year revenue growth and 25% year-over-year Adjusted EBITDA growth. Overall, I think our results speak for themselves."  

  • U.S. Cellular to launch LTE roaming in next 60-90 days:  CEO Ken Meyers said U.S. Cellular has completed its first LTE roaming agreement, though he declined to reveal the carrier partner. He said the companies are in the implementation phase of the deal and the respective engineering teams of the companies are working together. U.S. Cellular customers will be able to start benefiting from expanded LTE roaming in the next 60 to 90s days, he said. The partner is likely a Tier 1 carrier, so U.S. Cellular customers will get access to a more robust and nationwide LTE network. Meyers said he expects U.S. Cellular customers to see benefits more than he expects U.S. Cellular to reap inbound roaming revenue.  Meyers said that the agreement is the first of multiple LTE roaming deals the company is working on.


IHS-Infonetics: MSOs Plan Massive DOCSIS Deployments; Shift to Remote/Distributed Access & HFC Optical Nodes

1. IHS-Infonetics conducted in-depth interviews with cable operators (MSOs) across the globe that collectively control 87 percent of the world’s cable capex and found that 42 percent of them plan to deploy a distributed access architecture (DAA) by 2017.

In the study, CCAP, DOCSIS 3.1, and Distributed Access Strategies and Vendor Leadership: Global Cable Operator Survey, respondent operators say their primary choices for distributed access are R-PHY, R-MACPHY and R-CCAP.


  • The operational benefits cable operators are reaping from moving from CMTS (cable modem termination system) to CCAP (converged cable access platform) are just the first step in a long-term transition to distributing data processing capabilities throughout the network
  • Survey respondents, on average, say that about 1/3 of their residential subscribers will be passed by DOCSIS 3.1 (CableLabs spec) enabled headends by April 2017
  • By 2017, nearly half of respondents will have return path (upstream) frequencies of 86–100MHz, while a quarter will have 101–200MHz of return path spectrum


“Cable operators are clearly committed to both DOCSIS 3.1 and distributed access architectures to increase bandwidth in their access networks. Though there is no consensus yet on which distributed access technology most will use, there’s no question they will distribute some portion of the DOCSIS layer to their optical nodes,” said Jeff Heynen, research director for broadband access and pay TV at IHS.


●    Remote PHY (R-PHY): in this scenario, the entire DOCSIS PHY modulation is moved into the node while the MAC layer remains in the headend.

●    R-CMTS: in this scenario, the DOCSIS MAC and PHY are removed from the headend and placed in the node 

●    R-CCAP: in this scenario, the DOCSIS MAC and PHY and video QAM capabilities are removed from the headend and placed in the node. 





The 29-page IHS Infonetics study, led by IHS analyst Jeff Heynen, focuses on DOCSIS 3.1, converged cable access platforms (CCAPs) and distributed access architectures, and how and when cable operators will deploy these technologies and architectures to improve their broadband and IP service offerings over the next 2 years.

The study includes operator ratings of CCAP and distributed access equipment suppliers (Arris, Casa Systems, Cisco, Gainspeed, Harmonic, Huawei and Pace/Aurora) on 9 criteria.  Note from Jeff Heynen:  "These are really the primary suppliers of node-based products. There are a number of Chinese ODMs. But they are really only present in China proper."

2.  In a related report, IHS-Infonetics forecasts global hybrid fiber-coaxial (HFC) optical node shipments to more than double in the 5 years from 2014 to 2019, jumping from 92,000 to 200,000. Driving the boost are cable operators upgrading their networks with optical fiber cable, taking advantage of its high-bandwidth, low-noise, low-interference characteristics to deliver broadband video, data and voice services to homes and businesses.

“Optical nodes have rapidly become important platforms for cable operators to grow their broadband capabilities. By way of increased node splitting today for increased bandwidth and a transition to distributed access in the coming years, optical nodes will see significant unit growth and innovation,” said Jeff Heynen, research director for broadband access and pay TV at IHS.


  • Globally, HFC optical node revenue reached $356 million in 2014, up 14 percent from 2013
  • In 2014, 80 percent of worldwide optical node revenue came from digital return nodes, and 15 percent from analog return nodes
  • By 2019, IHS expects 35 percent of new physical nodes to be remote CCAP devices, 27 percent to be R-PHY units, and 23 percent to be traditional digital return nodes
  • Arris led optical node global revenue and physical node unit shipments for the full-year 2014



The 23-page IHS Infonetics HFC Optical Nodes market share and forecast report provides worldwide and regional market size, vendor market share, forecasts through 2019 and in-depth analysis for hybrid fiber-coaxial optical nodes. The annual market research service tracks physical node units, logical node segments and revenue for optical node types including analog and digital return, and remote PHY, converged cable access platform (CCAP) and cable modem termination system (CMTS).


To buy Infonetics reports, visit:

My perspective on ONS 2015, SDN & Open Networking

The 2015 Open Networking Summit (ONS) was hosted on June 14th to June 18th in Silicon Valley. It featured a rich set of speakers, open networking panels and a wide audience base comprised of network service providers, network hardware and software vendors, web giants and the academia.  ONS 2015 was the first of the fifth annual summit that I participated in.  To this onlooker it provided an insight into the future direction of networking.   The conference was a showcase of the solutions and challenges in achieving the goal of Software Defined Networking (SDN): to make the network programmable.    

As Chair of the IEEE Communications Society Santa Clara Valley (SCV) chapter, I've had the opportunity to host several technical sessions on Open Networking and also track the rapid pace of change towards SDN.  I am thrilled and enthused by the changes that SDN can provide.  It opens up significant opportunities for new and existing players.   However, I am equally skeptical of when and how SDN will become a mainstream technology, available to any enterprise data center or any end network consumer.  

The basic idea of Software Defined Networking is to make the network user programmable. Sounds simple? It depends on how one defines the network:  

  • Is it a home network, enterprise Local Area Network (LAN) network, ISP / telco / carrier network, a web giant network, Cloud Service Provider Network or a Wide Area Network (WAN)?
  • Is it a private network (located inside an enterprise and accessible only to an internal audience) or a public network (located on a premise not owned by the enterprise)?
  • Is it a physical network (network functionality achieved using dedicated hardware) or a virtual network (network functionality achieved by using software and white box hardware)?


Each network type has it’s own set of solution pieces offered by multiple vendors, which consist of hardware and software components that are provisioned and maintained by a Service Provider (data center, telco WAN, enterprise/campus, cloud computing/storage, etc).  Each network also has its own set of operational requirements.  There are a wide range of issues and concerns, including: security, availability, provisioning, power, cost and serviceability.   

The Open Network Foundation (ONF), which is progressing Open Flow based SDN, has a herculean task of bringing all these pieces under a single umbrella. Achieving SDN in an an “open,” “vendor agnostic” and “inter-operable” way is a challenge the purist can compare to finding extra terrestrial life.

Google’s Fellow and Technical Lead for Networking Amin Vahdat was an impressive keynote speaker at ONS2015.   For the first time in company's history, Amin disclosed Google's internal Data Center Network.  It’s design is based on the principles of Software Defined Networking, leverages CLOS topology, uses merchant silicon and has a single central administrative domain.  A few statistics that are indicative of this massive network are that it handles 3.5 billion search results per day and has 300 hours of video uploaded every minute!  Let’s pause for a minute and extrapolate, at roughly 5MB bandwidth consumption per minute for a 480p video - it translates into about 50 Petabytes of network traffic to watch the video content uploaded over a period of year (18000 years of uninterrupted viewing content stored and generated every year).

Microsoft’s Mark Russinovich, the CTO of the company's Azure public cloud was also a keynote speaker.  He talked about how Microsoft has embraced SDN into the Azure wide area network.  That network can host millions of compute instances, and has exabyte scale storage and a Petabit capacity (Pbps) network.

Note that both Microsoft and Google are competing with Amazon’s AWS (Amazon Web Services) - a cloud-compute service provider platform.

Given the scale out requirement to handle the data generated by the human race today, one thing is clear: SDN is not an option - it is the solution.  That's because large networks that have to rapidly increase the number of users and the aggregate data capacity  (e.g. Amazon's AWS, Microsoft Azure public cloud, Google's customer facing and backbone network, NTT and AT&T WANs, etc) require a software based approach with centralized domain specific control to scale out. The traditional hop by hop routing with expensive, closed, proprietary routers won't make the grade.

AT&T’s SVP John Donovan, was another keynote speaker. He highlighted the journey of transformation which AT&T is pursuing with Open Networking, SDN and Open Source software.  AT&T is on a grandiose mission to replace the traditional telephony network, based on Time Division Multiplexing (TDM)  to an all Ethernet network by 2020. 


The ONS2015 was spread over a week and had several panels on various important topics of SDN Adoption, Use Cases, Experiences, Hot Startups & VC investments, SDN WAN, Network Functions Virtualization (NFV), SDN for Optical Networks, OpenStack. The ONS2015 also had an expo floor, comprised of sponsor solution and demo booths from various companies like NEC, ADVA, AT&T, Dell, Brocade, Huawei, Cisco, and Broadcom.

There is a rapid pace of technology advancement, tremendous amount of energy and resources are being invested in this "second life of networking”. One pundit called it "a new epoch."  While there is market fragmentation and chaos, I see that as a positive sign.  The industry is moving forward, asking new questions, facing new challenges.  Consolidation is far ahead.  Let’s continue to build and play by the ONF vision to build, promote and adopt SDN through open standards and open source software development. 

I will close by a quote from Kitty Pang (Network Architect, Alibaba).  It is bold and provocative, yet real: 

"We want to run faster and faster.  It does not matter if it’s hardware or software, open or closed, we choose low cost and high efficiency."


Watch an insightful interview, where Alan Weissberger talks to ONF’s Dan Pitt on ONF’s path towards Open SDN:


Editor's Note:

Saurabh Sureka is the Chair of IEEE ComSoc Santa Clara Valley (SCV), which is by far the leading ComSoc chapter in the world in terms of both membership and technical programs.   He joined the leadership team in 2011 as Treasurer and diligently continued to volunteer each year since then as a ComSocSCV officer.   Saurabh is a Sr Product Manager at Emulex (now Avago Technologies) in San Jose, CA.

Related articles: 

Highlights of 2015 Open Network Summit (ONS) & Key Take-Aways

Pica8 Open Networking OS/Protocol Stacks on Bare Metal Switches



FCC Approves AT&T-Direc-TV; Imposes Conditions to Improve Broadband Competition?

After almost one year of regulatory review, AT&T closed its $49 billion acquisition of DirecTV which makes it the largest U.S. pay-TV company.  As expected, the Federal Communications Commission announced on Friday that it's approved AT&T's merger with DirecTV, attaching conditions intended to address the potential harms of the merger. Earlier this week, the U.S. Justice Department announced that it would not challenge the acquisition. 

“The conditions also ensure that the benefits of the merger will be realized,” the FCC said in a news release.  Federal regulators were reviewing the deal to determine whether it would serve the public interest or stifle competition.  Those regulators have said they're more worried about providing choice in Internet access and new, online video options than they are about concentration in the declining pay TV business (go-go cord-cutters!).

 -->We strongly feel it's the latter- less choice of providers leads to less competition and ultimately higher prices!

“We’re now a fundamentally different company,” AT&T Chief Executive  Randall Stephenson in a press release.   The company said it will serve more than 26 million U.S. customers and more than 19 million in Latin America, making it the world’s biggest pay-TV company.

[AT&T reported its earnings on Thursday.  Profits dropped 14% to $3.04 billion.  The company said integration related expenses from prior deals weighed on the results.  Revenue edged up to $33.02 billion- an increase of only 1.4%.  The number of new mainstream wireless subscribers fell by 60%.  AT&T's annual revenue growth since 2007 has averaged a miniscule 1.6%, according to data from FactSet.]  

AT&T Chief Strategy Officer John Stankey will be chief executive of a new division called AT&T Entertainment & Internet Services, which includes DirecTV and the division that also includes AT&T’s broadband and video business.  As the biggest pay-TV provider, AT&T could have more bargaining power with content companies.

“We are more confident than ever about the opportunity this transaction brings,” AT&T Chief Financial Officer John Stephens said on a conference call Thursday.

“We’ll now be able to meet consumers’ future entertainment preferences, whether they want traditional TV service with premier programming, their favorite content on a mobile device, or video streamed over the Internet to any screen,” Randall Stephenson, chairman and chief executive of AT&T, said in a statement.


Conditions Imposed on the 2nd "new AT&T":  

Note: the 1st "new AT&T" was when SBC acquired AT&T in 2006 and kept the AT&T name/

Approval of the deal came with a number of conditions, including some aimed at introducing more competition into the broadband Internet market, an issue emphasized by FCC commissioner Tom Wheeler in his comments earlier this week.The FCC is requiring AT&T to expand its high-speed, fiber-optic broadband Internet service to 12.5 million customer locations and eligible schools and libraries. That’s about 10 times its current size. The FCC said this addresses the concern that the merger would eliminate one choice for television service in the areas where AT&T and DirecTV previously competed. By expanding Internet service, the commission said, consumers will have more options to use services that rely on broadband to deliver video, such as Netflix, Amazon and Hulu.  AT&T also will be required to offer broadband services to people with low incomes at discounted rates.

The company also will be required to submit its carrier inter-connection agreements for review by the FCC.  Those agreements include "paid peering," which allow a video streaming company like Netflix to pay a fee to a distributor, like Comcast or AT&T, for better service, when they create a lot of traffic for the network. The commission said that the condition recognized the importance of those agreements to online video service and said that it would monitor them to make sure that AT&T would not deny or impede access to its networks in anti-competitive ways.

The conditions remain in effect for four years after the merger closes. The FCC also required AT&T to retain an internal compliance officer and an independent, external compliance officer to make sure that the company abides by the deal conditions.  

A serious concern about the deal is that AT&T is the only major Internet Service Provider whose customers face “data caps” for wireline broadband Internet access1.  The merger could increase the incentive of AT&T to deploy such usage-based pricing to limit access to online video in favor of its own traditional television service. As a condition of the deal, regulators forbade AT&T from deploying discriminatory practices that would disadvantage online video services.

Note 1.  AT&T Data Caps: "Residential AT&T High Speed Internet service includes 150 gigabytes (GB) of data each billing period, and most residential AT&T U-verse High Speed Internet service (up to 75 Mbps) includes 250GB of data each billing period."


Public Interest Groups Weigh In:

Some public interest groups, though, were disappointed. “I thought after the Comcast-Time Warner Cable deal that maybe the commission was going to travel down a little different road in consolidation and begin to say no to some of these deals,” said Michael Copps, a former Democratic member of the FCC and a special adviser to the Common Cause public interest group.

“What they are basically saying is you have to treat everybody like you treat yourself, and so I think that is probably the most important protection against anticompetitive practices,” said Gene Kimmelman, the chief executive of Public Knowledge, a consumer advocacy group, and a former antitrust official at the Justice Department.


More Media Mergers Ahead?

The combination of AT&T, one of the country’s two largest wireless/wire-line telco and Internet Service Providers (via AT&T-Yahoo), and DirecTV, the country’s largest satellite TV provider, is the biggest media merger this year and will create the country’s largest television distributor with about 26 million subscribers, surpassing Comcast, the current leader.

“The fact that this deal closed with probably pretty reasonable conditions gives a little bit more confidence that Charter and Time Warner Cable would close, and maybe down the road opens the door for other deals,” said Amy Yong, a media analyst with Macquarie Group.


IHS-Infonetics: True 4G (LTE Advanced) is Finally Happening; SPs What's 5G?

IHS-Infonetics today released excerpts from its 2015 IHS Infonetics 4G and 5G Strategies and Vendor Leadership: Global Service Provider Survey, for which operators were interviewed about their LTE network deployment plans, challenges and service offerings. Half of respondents participating in the study say they have already deployed LTE-Advanced (LTE-A) in their LTE networks.


  • Inter-band carrier aggregation is the most common and very first LTE-Advanced feature deployed by respondent operators.
  • Commercial voice over LTE (VoLTE) service is taking off slowly and ramping this year and next.
  • 4G network functions virtualization (NFV) migration won't happen any time soon because the bulk of LTE networks are brand new and, therefore, mobile operators are not ready to undertake migration that soon.
  • Ericsson, Huawei and Nokia (in alphabetical order) are perceived by survey respondents as the top LTE equipment manufacturers.


Analyst Comments:

"We are slowly but surely moving to true 4G, and that's good news. However, most users already believe they are on 4G, and that's the bad news because the experience is far from consistent and is falling short of expectations. How many times does your smartphone display LTE or 4G and you still see the infamous spinning wheel?" said Stéphane Téral, research director for mobile infrastructure and carrier economics at IHS. 

"The 5G debate has started with great fanfare, hype and confusion, but little substance about what it is exactly and what it is not. For now, the mindset is still locked into mobile broadband as we know it with LTE, so it's good that the ITU has just stepped in to define 5G in its brand new IMT-2020," Téral said. 



Author's Notes:

1.  What survey respondents think 5G will be at this time is anyone's guess.  That's because ITU-R hasn't even finalized the 5G vision or architecture recommendations.  Please refer to this post for the status of all the 5G work in progress:

2.  One huge problem for "true 4G" users is there is no set of minimum service requirements a wireless carrier has to implement to claim compliance with LTE Advanced.  It seems that a higher data rate than LTE along with Carrier Aggregation is what most wireless operators are implementing or planning.  See Appendix for LTE Advanced Requirements.

In a July 23rd email, Stephane Teral corroborates the above problem:

Hi Alan,  I agree and this is exactly what’s looming for 5G if no one, including the ITU, does not step in to call for the clear cut! At this point, everyone implementing carrier aggregation can claim true 4G (IMT-Advanced), which is much better than saying LTE is 4G when it is defined in the IMT2000 as a 3G technology!

Best wishes,



 For the 38-page 2015 4G and 5G Strategies and Vendor Leadership: Global Service Provider Survey, IHS interviewed purchase-decision makers at 22 mobile, incumbent, competitive and cable operators from EMEA, Asia, North America and Latin America. The study covers LTE network build-out plans; challenges and drivers; migration scenarios; LTE features, services and suppliers; and operator ratings of LTE manufacturers (Alcatel-Lucent, Cisco, Ericsson, Huawei, Nokia, Samsung, ZTE) on 9 buying criteria.

The service providers participating in the study represent about one-third of the world's telecom capex and revenue.

To purchase the report, please visit:



Mobile Services Market Dragged by Europe Again:

LTE Peaking at $6 Billion a Quarter - Not Enough to Offset 2G/3G Decline:

Operators Spent $67B Outsourcing Network Tasks to Equipment Vendors in 2014:

Telecom Carrier Spending Entering New Era Marked by Diverse Regional Trends:

Mobile Operators Using EDGE, HSPA+ to Improve User Experience on Road to LTE:


Appendix:  LTE Advanced Requirements:

General Requirements

  • LTE-Advanced is an evolution of LTE
  • LTE-Advanced shall meet or exceed IMT-Advanced requirements within the ITU-R time plan
  • Extended LTE-Advanced targets are adoptedSystem

System Performance Requirements

  • Peak data rate

 - 1 Gbpsdata rate will be achieved by 4-by-4 MIMO and transmission bandwidth wider than approximately 70 MHz

  • Peak spectrum efficiency

 - DL: Rel. 8 LTE satisfies IMT-Advanced requirement
 - UL: Need to double from Release 8 to satisfy IMT-Advanced requirement

  • Capacity and cell-edge user throughput

 - Target for LTE-Advanced was set considering gain of 1.4 to 1.6 from Release 8 LTE performance


Other Important Requirements

  • Spectrum flexibility

 - Actual available spectra are different according to each region or country
 - In 3GPP, various deployment scenarios for spectrum allocation are being taken into consideration in feasibility study
 - Support for flexible deployment scenarios including downlink/uplink asymmetric bandwidth allocation for FDD and non‐contiguous spectrum allocationTotal

  • LTE-Advanced will be deployed as an evolution of LTE Release 8 and on new bands.
  • LTE-Advanced shall be backwards compatible with LTE Release 8 in the sense that

 - a LTE Release 8 terminal can work in an LTE-Advanced NW,
 - an LTE-Advanced terminal can work in an LTE Release 8 NW

  • Increased deployment of indoor eNBand HNB in LTE-Advanced.



FBR & Co: Verizon Wireless Market Comments + Verizon's Mobile Video Service & "skinny bundles"

From FBR's David Dixon:

"Verizon has a sustainable industry leading (wireless) network, an excellent spectrum position, and surging growth opportunities in mobile video, Internet of Things, and telematics."

"We expect T-Mobile US to continue affecting AT&T to a greater extent than it effects Verizon.  Sprint will remain weak as it delays spending amid a network strategy, driven by a weak balance sheet."

Editor's Note: Verizon is the first big telecom company to report its earnings for the 2nd quarter.  It has faced tougher competition as such rivals as  T-Mobile USInc. and Sprint Corp. have offered generous deals to subscribers to switch.


1.  Aside from being well positioned on spectrum for the macro network, how should investors assess the small cell opportunity as an alternative to more macro network spectrum going forward?

A change in the industry network engineering business model is underway toward using small cells on dedicated spectrum to manage more of the heavy lifting associated with data congestion. Verizon demonstrated this shift during the AWS3 auction: It modeled a lower-cost small cell network for Chicago and New York. We expect VZ CEO Lowell McAdam to manage this shift from the top down to mitigate execution risk due to cultural resistance from legacy outdoor RF design engineers, whose roles are at risk as the macro network is de-emphasized.

Enablers include the advent of LTE, increased spectrum supply across multiple spectrum bands including LTE licensed, unlicensed (e.g., 500 MHz of 5 GHz spectrum) and shared frequencies (e.g., 150 MHz of 3.5 GHz spectrum, amid a fundamental FCC spectrum policy shift from exclusive spectrum rights to usage-based spectrum rights, which should dramatically increase LTE spectrum utilization similarly to WiFi.  

Previously, outdoor small cells co-channeled with the macro network proved challenging:  While they can carry substantial load, they also destroy equivalent capacity on the macro network due to mis-coordination and interference. So the macro network carries less traffic, but still looks fully loaded. AT&T discovered this in its St. Louis trials that in part steered it toward buying $20 billion of AWS3 spectrum. However, the industry trend is toward LTE underlay networks, where small cells are put in other shared or unlicensed spectrum with supervision from and/or carrier aggregation with the macro network.

It still requires good coordination across all cells for this to work; while Verizon s initial proposals for 5 GHz are downlink only, we think uplink will also be used longer term because the uplink needs more spectrum resources for a given throughput and we are seeing higher uplink usage trends in the Asian enterprise segment and from Internet of Things (security cameras).

2.  Does Verizon have sufficient spectrum depth to drive revenue growth longer term? Or does it need to aggressively acquire spectrum in future spectrum auctions or in the secondary market (e.g., DISH)?

The short answer is yes. Verizon carries 80% of data traffic on 40% of its spectrum portfolio; its combined nationwide CDMA and LTE spectrum depth is 115 MHz, ranging from 88 MHz (Denver) to 127 MHz (NYC). We expect AWS3 capacity spectrum to be deployed in 2017/18.

Investors may not be crediting Verizon with potential to source more LTE spectrum from refarming of CDMA to LTE (22 MHz to 25 MHz) used today for CDMA data (22 MHz to 25 MHz). Critically, network performance data show Verizon network close to the required performance threshold for a VoLTE-only service, suggesting there is additional refarming potential for the 850 MHz band (25 MHz) used today for CDMA voice and text. This band is likely to be transitioned in 5 MHz x 5 MHz LTE slivers to run parallel with the expected linear (voluntary) ramp, versus exponential (forced) ramp in VoLTE service. More low-band spectrum is key for the surging IoT and M2M segments, which are proving to be more thirsty than "bursty.


Verizon's CFO Shammo talks up mobile-first video service 
Verizon Communications plans to offer a limited menu of content when it initially launches its mobile-first streaming video service late this summer, Chief Financial Officer Fran Shammo said after the telecom announced earnings Tuesday. The service is expected to use LTE Multicast technology. RCR Wireless News (7/21), CNET (7/21)

Shammo added that interest in Verizon’s FiOS skinny TV channel bundles exceeded the company’s expectations as one-third of gross customer additions chose the skinny bundle and some existing customers migrated to it as well. Verizon added a net 26,000 FiOS video subscribers in the three months ended June 30, and a net 72,000 subscribers to its FiOS Internet service.

Verizon's 2Q 2015 Earnings Report:


IHS-Infonetics: NFV Market to Grow More than 500% Through 2019; Alan Disagrees!

IHS-Infonetics released excerpts from its IHS Infonetics NFV Hardware, Software, and Services report, which forecasts the global network functions virtualization (NFV) hardware, software and services market to reach $11.6 billion in 2019, up from $2.3 billion in 2015.


  • Service providers are still early in the long-term, 10- to 15-year transformation to virtualized networks
  • Revenue from outsourced services for NFV projects is projected to grow at a 71% compound annual growth rate (CAGR) from 2014 to 2019
  • Revenue from software-only video content delivery network (CDN) functions for managing and distributing data is forecast by IHS to grow 30-fold from 2015 to 2019


Analyst Quote:

“NFV represents operators’ shift from a hardware focus to software focus, and our forecasts show this. We believe NFV software will comprise over 80 percent of the $11.6 billion total NFV revenue in 2019,” said Michael Howard, senior research director for carrier networks at IHS.   “The software is always a much larger investment than the server, storage and switch hardware, representing about $4 of every $5 spent on NFV,” Howard said.




The 2015 IHS Infonetics NFV Hardware, Software, and Services market research report tracks outsourced services for network functions virtualization (NFV) projects as well as service provider NFV hardware, including NFV infrastructure (NFVI) servers, storage and switches; and NFV software split out by service management and orchestration (NFV MANO) software and virtual network function (VNF) software, including virtual routers (vRouters) and the software-only functions of mobile core and EPC, IMS, PCRF and DPI, security, video content delivery networks (CDN), and other VNF software. The research service provides worldwide and regional market size, forecasts through 2019, in-depth analysis and trends.

To purchase the report, please


Watch analyst Michael Howard’s July 2015 webinar, SDN and NFV: Accelerating PoCs to Live Commercial Deployment, an event detailing how operators can validate VNFs, NFVI and network services to speed up commercial deployments. Log in to view:


Alan's DIssenting View of NFV:

There've been many times in the last three or four decades where network operators were wildly enthusiastic about a new technology, which never really gained marrket traction.   In the mid 1980's, ISDN was to replace the PSTN phone system and usher in new world of data communicaitons.  In the early 1990s it was SMDS (Switched Multi-Megabit Data Service) that never really saw the light of day.  In the mid 1990s, ATM was going to take over the world.  10 years later, many operators including SPRINT believed that WiMAX would be the 4G technology of choice.  Didn't happen!

Now we have major telecom carriers all excited over NFV with many forecasts of huge growth of that market.  We don't think that will happen in the next five years or maybe ever.  Here's what's missing from NFV:

  1. Implementable standards for exposed physical interfaces and APIs for virtual appliances1 implemented as software running on a generic/commodity compute server.
  2. Procedures and protocols for virtual appliances to communicate with  physical box (legacy, non-virtual) appliances already deployed in carrier networks (otherwise known as backward compatibility with the installed base).   An ETSI NFV Use Cases document states in section 7.4.  Coexistance of Virtualized and Non-Virtualized Network Functions:  "The communications with virtual network functions shall be based on standardized interfaces."  Yet those interfaces haven't been standardized yet.
  3. New security methods and procedures to isolate and quarantine compromised virtual appliances so that all appliances running on the same compute server are not locked down.
  4. Hardware assists, possibly NICs for compute servers, to guarantee latency and throughput of virtual appliances running on commodity compute servers.
  5. Standardized Management & Orchestration (MANO) software, including "service chaining," which will work with different virtual appliances from multiple software vendors. Standardized software interfaces (like APIs) will be required for this, perhaps as part of an open source MANO software package (from OPNFV?)
  6. Element Management Systems (EMSs) that are integrated into MANO.  For decades, physical box network vendors provided their own proprietary EMSs which configured, managed, monitored and controlled their equipment. EMSs need to be converted to software modules within MANO or accessible to MANO via software interfaces.
  7. The ability of MANO or other software to manage, control, schedule appliances/services, etc from BOTH virtual appliances and physical box appliances.  Note that both types of appliances will be used in telco central offices/data centers for many years.  
  8. Procedures for testing, monitoring, OA&M, fault isolation, repair & restoration, etc are urgently needed.
  9. Open Source NFV software, perhaps from the OPNFV consortium.  IMHO, this is the best hope for NFV being a commercial success/real market.
  10.  There's a lot of hype about virtualizing the LTE Evolved Packet Core (EPC) via NFV but that is nonsense because operators are not going to put in a new infrastructure after recently spending lot of money to install EPC equipment and software.  Further, there are no standards for vEPC.   Yet we see many blog posts/article that say the time is now for vEPC.  Here's one of many NFV -vEPC hype to Pluto blog posts:              


Note 1.  Examples of virtual appliances include: session border controllers, load balancers, deep packet inspection agents, firewalls, intrusion detection devices, and WAN accelerators.


Here's an interesting quote from a Spirent white paper on NFV:

"Traditionally, the burden of validating the core functions is shouldered by the network equipment (box) vendors. Today’s NFV landscape shifts some of this burden to operators deploying NFV topologies while not absolving the need for network vendors to validate NFV software within different hardware (compute platforms, physical switching tiers) and software infrastructures (hypervisor, holistic cloud stacks public clouds, etc)."

Tom Nolle on Fixing NFV:

In a New IP blog post, Tom Nolle wrote:

"What's needed in NFV is something like "NFVI plug-and-play," meaning that any hardware that can be used to provide hosting or connectivity for virtual functions should be capable of being plugged into MANO and supporting deployment and management."

Nolle goes on to elaborate on fixes for CAPEX, OPEX and business case.  He concludes with this remark:

"If the ETSI NFV ISG, or Open Platform for NFV Project Inc. , or the New IP Agency, are serious about moving NFV optimally forward, I'd suggest this is the way to start."