Archive | Service provider RSS feed for this section

Employing AI to Enhance Returns on 5G Network Investments

13 Sep

Wireless 20/20 believes that AI will play a crucial role in helping operators to maximize returns on their 5G network investments. AI will open exciting opportunities for the mobile communications sector to proactively manage the costs of deploying and maintaining new 5G networks while helping to create a more personal approach for customers.

In October and November of 2018 Ericsson conducted an AI survey, showcasing service providers that have adopted AI to manage the costs of deploying and maintaining mobile communications networks while helping to create a more personal approach for managing customer relationships customers. The Ericsson Report concludes that that more than half of mobile operators—a total of 53%—expect to have adopted AI within their 5G networks by the end of 2020.

AI in 5G Network Planning and Performance Management

5G is expected to cover more than 40% of the world’s population, and total mobile data traffic is predicted to have increased by a factor of 5 by 2024. With the advent of 5G, service providers are making huge investments in their networks to enable the new use cases that 5G offers. Ericsson research reveals that wireless service providers around the world are presently at various stages on their journey with AI and 5G. Early adopters of AI among service providers will undoubtedly gain an advantage, as they will be well–placed to deal with new challenges that result from the proliferation of additional devices following the introduction of 5G. This is because the advent of 5G will make network topologies relatively complex, with small cells and new antennas making usage patterns more difficult for humans alone to predict, and current radio propagation models becoming more complex to compute as a result of new radio spectrum bands, denser topologies, Massive MIMO, and beamforming.

The Ericsson Report also reveals that AI is presently facilitating improvements ranging from simplifying network evolution to improving performance across existing networks. Most service providers are at the stage of testing AI, with 48% focusing on AI to reduce capital expenditures. Service providers believe the highest potential return from AI adoption will be in network planning (70%) while 64% intend to maximize their returns by focusing their AI adoption efforts on network performance management. The highest current focus of AI initiatives among wireless service providers worldwide is in service quality management (17%) and operational cost savings (16%). A further 41% are focusing on using AI for optimizing network performance, and 35% for new revenue streams. AI and machine learning will enable wireless network vendors to quickly process raw data and deliver analytical outcomes to help operators more rapidly recoup their 5G network investments.

AI Will Be Vital to 5G Customer Service

For service providers, AI offers opportunities to build solutions jointly with infrastructure providers, with a common goal to more effectively manage complexity and optimize network performance. Service providers around the world are observing improved reliability for customers as the area in which AI is currently having the greatest impact upon core network activities. The Ericsson survey demonstrates that AI is creating both benefits and challenges for service providers at the advent of 5G. Enhancing customer experience was identified by 55% of service providers as a key area where AI is presently having the greatest impact within core network activities. In addition, 68% of survey respondents highlighted enhancing customer service as a business and operational objective over the next 3 years. A further 72% agreed strongly that AI will be important in enabling monetization of new network technologies and providing a better service to customers. A wide range of operators are already beginning to enhance big data with AI to automate customer service with intelligent assistants and chatbots.

Many service providers are already concluding successful trials on using AI in their networks. Only 12% feel they have a detailed knowledge of AI’s application. However, 49% considered themselves to have a fairly detailed knowledge of AI application. More than half expect AI to be adopted in their networks before the end of 2020 (a total of 53% globally) and there is a general expectation (55% globally) that the benefits will be evident within 1 to 2 years.

AI to Deliver Optimal 5G User Experience

Ericsson is convinced that AI offers the best opportunity to achieve the high levels of automation necessary to optimize and manage the complexity of 5G system performance, allowing them a shift from managing networks to managing services. As 5G-enabled technologies develop, operators will need AI to augment the human capabilities to improve efficiency and manage their OPEX. Ericsson has introduced engineering solutions that combine AI, machine learning, and human ingenuity to enable networks to self-learn, self-optimize and deliver optimal user experience, allowing operator customers to capitalize on the opportunities of 5G. Ericsson believes that AI and machine learning will be crucial to the evolution of 5G network automation, IoT and industrial digitalization.

AI and 5G in Europe

Vodafone is one Ericsson customer that is leading the industry in using AI in radio networks, because of the pioneering work between Ericsson and the operator’s Networks Centre of Excellence. Vodafone and Ericsson are collaborating to develop advanced AI and Machine Learning algorithms. One use case is for Vodafone to improve MIMO energy management at radio sites by putting radio transmitters into power-saving Sleep Mode when traffic falls below certain levels and then re-activate them automatically when traffic surges. Vodafone has deployed 5G in seven UK cities, including London. Backed by its largest ever capital investment in partnership with Ericsson, Vodafone is enabling Londoners to access the new ultra-fast 5G network without any limits based on new unlimited data plans. Vodafone will provide comprehensive 5G coverage in London, leveraging the latest Ericsson Radio System portfolio, including the latest Baseband 6630 and Massive MIMO 6488 products to enable 5G using the 3.5GHz frequency. Combined with LTE, this will achieve speeds up to 10 times faster than 4G for 5G users with much lower latency. Vodafone Spain recently launched 5G services in three cities operating in the 3.7 GHz band utilizing Ericsson products and solutions. Vodafone and Ericsson have also launched a commercial 5G network in Germany with the goal of bringing 5G to 20 million people in over 20 cities by the end of 2021. Vodafone Business and IBM will also supply enterprise customers with managed services in the areas of cloud and hosting and will work together to build and deliver solutions in areas like AI, cloud, 5G, IoT, and software-defined networking.

AI and 5G in the US

AT&T and Tech Mahindra are collaborating to build an open source artificial intelligence (AI) platform called Acumos, which will make it easy to build, share, and deploy AI applications. The Acumos AI Marketplace is an extensible framework for machine learning solutions which provides the capability to edit, integrate, compose, package, train, and deploy AI microservices. By getting developers and businesses to collaborate effectively, AT&T will industrialize the deployment of AI at enterprises to deliver tangible value and solve real business problems. AT&T has used the model of moving its own technology into the open source community to engage developers and accelerate the development of the platform. AT&T is collaborating with Tech Mahindra and to make AI simpler to improve adoption and help enable enterprises apply AI to reimagine business models, unlock the potential of data and drive business outcomes. Tech Mahindra has now expanded this strategic collaboration to assume management of many of the applications which support AT&T’s network and shared systems. The goal is to accelerate AT&T’s IT network application, shared systems modernization, and movement to the cloud. This partnership should significantly boost AT&T’s 5G time-to-market and simultaneously reduce their cost of ownership by automating aspects of their network lifecycle. Manish Vyas, President, Communications, Media and Entertainment Business and the CEO, Network, Tech Mahindra will present the Acumos Project on October 25 in Track 12 at AI World.

T-Mobile is also leveraging AI and machine learning to completely overhaul and accelerate the automation of its customer service operations in the US. T-Mobile is using the predictive capabilities of AI and machine learning to augment human abilities and reshape its customer service. T-Mobile customers immediately connect with a live customer service agent that knows them, rather than talking to an IVR or chatbot. With the help of AI, these customer service agents can quickly access the information most salient to customer needs. These AI-driven customer care initiatives will be critical as T-Mobile prepares to deliver nationwide 5G using a mix of wireless spectrum. T-Mobile plans to introduce standalone 5G in 2020 and recently accomplished the world’s first standalone 5G data session on a multi-vendor 5G next generation radio access and core network—and the first standalone 5G data session of any kind in North America. However, T-Mobile has placed some of its 5G network deployment efforts on hold amid regulatory delays in its pending Sprint merger.

Verizon plans to invest between $17 billion and $18 billion in capital expenditure as it builds out 5G networks and launches 5G services in 30 markets based on millimeter wave spectrum in 2019. During Track 12 at AI World, Verizon speakers will discuss efforts to leverage 5G, AI and Mobile Edge Computing, with the aim to have some commercial services on this infrastructure by late 2019. By installing IT and network-processing resources in data centers at the network edge, instead of in the centralized facilities where they are normally found, operators could shorten the journey for a data signal and reduce latency. Verizon is confident it will be able to cut latency by at least 80% through investment in 5G technology and the rollout of new “edge” architecture. Verizon is testing a cloud gaming platform and this latency reduction could lead to new service opportunities in areas such as virtual-reality gaming services and self-driving cars.

Verizon is also enhancing its portfolio of managed services with an AI-powered toolkit for improving 5G customer experience outcomes. Verizon has made a large investment in AI and machine learning technologies and uses advanced predictive analytics algorithms to deliver “Digital Customer Experience” offerings for businesses. Verizon’s new Digital Customer Experience platform combines four AI-powered components to improve customer support outcomes: virtual agent, live agent, knowledge assist, and social engagement. Verizon believes the use of AI in customer service is likely to increase in the near future and is integrating AI into its existing customer support pipeline, providing virtual assistance 24/7 via social media, chat services, email, text message, or phone, with support experiences based on past interactions. Verizon’s Virtual Agent platform incorporates AI to solve customer challenges on the spot and escalates users to human support agents when presented with a situation in which it is unable to help. The Knowledge Assist component combines authoring tools with machine learning to provide relevant answers and guidance for agents.

Ericsson Survey Methodology

This Ericsson report provides unique insights into the increasing need for relevant data about how service providers plan to integrate AI in to their 5G networks, based on a global comparison of high-level business objectives. This Ericsson survey was based on telephone interviews conducted by Coleman Parkes Research with 165 senior executives from 132 mobile communications service providers globally. The respondents were segmented into six categories based on function: Chief Technology Officers (CTOs), Chief Operating Officers (COOs), Chief Information Officers (CIOs), Chief Marketing Officers (CMOs), Chief Financial Officers (CFOs), and Line of Business (LOB) managers. Mapping the response of 165 executives from 132 mobile operators across the globe, the report provides valuable insights about the reasoning and expectations of using AI applications across 5G networks.

machine-learning-and-ai-aw-screen (Ericsson report)


Dutch virtual mobile market drops to 7.5 mln Sims

24 Dec
The number of mobile customers with virtual brands in the Netherlands fell by 15 percent in the six months to September 2015, after KPN’s brand Hi exited the market. The remaining 7.5 million customers at virtual operators still accounted for around 37 percent of the total Dutch mobile market, according to Telecompaper’s latest Dutch Mobile Virtual Operators Market report. Of the total, around 3.3 million were customers at operator-owned virtual brands, such as hollandsnieuwe or Ben, while the independent MVNOs counted 4.2 million Sims, or about 20 percent of the total Dutch mobile market.With the exit of Hi, KPN’s other brand Telfort became the biggest VO in the Dutch market. In terms of number of customers, the next four places were taken by Lebara, Tele2, Lycamobile and Vodafone’s hollandsnieuwe. These top five brands together accounted for 60 percent of the total virtual mobile market, an increase compared to six months earlier due to growth at all the players except Telfort.

While Tele2 has launched its own mobile network, it still relies to a certain extent on its MVNO agreement to serve customers, so it is included in Telecompaper’s VO figures for Q3 2015. With the next Virtual Operators Market report in Q1 2016, Tele2 will be excluded, as it is expected to have completed its nationwide network roll-out by then.

The ‘no-frills’ segment, offering basic services at relatively low prices, remains the largest segment in the VO market, accounting for more than 40 percent of customers. It grew by 3 percent compared to Q1 2015. All of the players in this segment, except Telfort and Simyo, managed to increase their customer numbers in the latest period, with Simpel and Youfone gaining the most. The next largest segment, taking about a third of the total VO market, remains the ethnic/international players, with Lebara taking the lead. The fixed operators, the third-largest segment, are also expanding in the mobile market, mainly due to growth at Tele2 and Ziggo.

Telecompaper’s latest 2015-Q3 report on Dutch VOs show that the total number of active virtual brands increased to 87 at the end of September 2015. Compared to Q1 2015, the total includes eight new entrants and three exits, while we also added two previously unknown players, which have been active a while. The market remains quite crowded, and several, particularly business providers continue to launch new initiatives, including mobile services as an add-on to their existing services. The number of launches as a standalone mobile-only solution was limited to one, the family-focused MVNO OpenMobile in May 2015. Other entrants included business fixed fibre player Fieber and six other business service providers.

Telecompaper expects more new entrants in the VO market, but also several exits, particularly amongst the independent MVNOs that only offer basic mobile services and depend on mobile as their main source of income. They will need to differentiate themselves better, as offering just a ‘me-too’ service is not a sustainable strategy, according to the researchers. The intense competition in the telecom market and changing interests among consumers are also driving MVNOs to launch value-added and OTT services, in order to limit churn and boost revenues, as recently seen at Lebara.

Much of the success of MVNOs depends on their wholesale agreement for network access. With Tele2 becoming the fourth mobile network operator, Telecompaper expects new opportunities to emerge for VOs. Tele2 is expected to actively seek wholesale customers on its network, as it works to raise capacity utilisation and recoup some of its network investment and operating costs. This may present a new chance for virtual players looking to gain access to 4G data services.


The Future of Telecom and Digital Services: Predictions for 2016

16 Dec

Today, telecom companies are undergoing a transformation in order to adapt their business models in a way that better engages the modern digital consumer. In 2015 we saw a great deal of shake-up in the media landscape, including a number of mergers and acquisitions and a continued rise of alternative service offerings, all pointing to an uncertain future for the market.As we evaluate what the future holds, and service providers shift their focus toward the customer experience to better engage the next generation consumer, there are a few trends that I predict will drive telecom transformation in the new year:

Service providers will try to appeal to cord cutters with new subscriptions offerings, including “skinny bundles” 

While the majority (83 percent) of American still pay a monthly cable fee, there is a growing number of consumers who are looking for options outside the standard cable and mobile packages. These “cord cutters,” as they have been dubbed by the media, want greater variety and personalization when it comes to their subscription offerings. In 2016, service providers need to focus on offering unique and customizable skinny bundles. In comparison to traditional offerings, these slimmed-down bundles would need to offer an a-la-carte feel that is attractive to consumers today, especially millennial audiences. By offering this more personalized approach, service providers can not only attract subscribers, but also begin to better engage today’s viewing audiences by providing a more customized experience.

Implementation of new distribution models and a focus on the Quality of Experience will be critical to engaging digital consumers 

The Quality of Experience will be a crucial differentiator for digital service providers in 2016. With video traffic growing to make up 80 percent of all consumer internet traffic by 2019, creating a seamless and high quality experience across devices will be critical for service providers in the new year. Digital consumers want to have access to their favorite shows from anywhere and at anytime – so they can start watching their favorite show on their tablet from the road, and finish it from their couch when they get home. Furthermore, personalization is imperative not only in the bundle offering, but also in the viewing experience. By implementing customized recommendation engines that tell viewers what they should view next based on their history, or providing integration with social media to encourage active viewing and sharing, service providers can enhance the viewing experience to make it easy, seamless and “sticky.”

Providers will increasingly innovate around their portfolio of offerings, moving beyond television viewing to attract consumers in new ways 

As digital services continue to become a major focus for service providers, my third prediction is that we will increasingly see telecom providers dipping their toes into new, and potentially unexpected, pools. Being the network provider is a huge asset for cable companies – whether wired or wireless, it gives them an almost limitless market for innovation, not only around pricing, but on potential future bundles as well.

The future service provider has the opportunity to introduce a portfolio of offerings. By exploring creative pricing bundles when integrating services such as mobile or integrating aspects of the connected home, cable companies can take advantage of data networks to fuel future bundles.

In looking to 2016, telecom providers face a number of challenges in defining their future role. However, there is also a great deal of opportunity to engage digital consumers by focusing on personalization in subscription offerings, the quality of experience across devices and exploring new innovative offerings.


Telcos urged to innovate data tariffs to counter OTT threat

24 Nov

At LTE North America in Dallas this week, a panel featuring Wireless 2020, IBM, Allot Communications and US-based MVNO FreedomPop urged the service provider community to develop more flexible and tailored data bundles for customers.

During a discussion focused on future revenue generation opportunities, conversation moved towards the current landscape which is seeing telcos lose ground to over the top content providers. Ken Jackson from IBM Now Factory believes that moving to specific and tailored, content-based services is a feasible opportunity for operators to monetise services in new ways.

“Let’s look at the example of a service called NFL Now, a friend of mine can watch as much NFL as he wants, and doesn’t pay for the data he uses, he pays to watch football,” he said. “We have to return the value proposition to the customer, find out what it is they want to do, and offer it to them. Make everything customer centric, and orchestrate your business around what they need”

Haig Sarkissian from Wireless 2020 concurred with Jackson, and urged the service provider community to move away from basic all-you-can-eat data plans.

“5% of the heaviest users consume 40-50% of the network, and they pay the same as the majority,” he said. “That’s why unlimited data plans are unfair, because the majority are subsidising the data usage of the minority. Operators find that this is no longer scalable because if you eat more you have to invest more into capacity. I don’t see it changing any time in the future, unless there’s a fair way of using these “dumb pipe” plans. Is there hope for SP differentiation outside of these unlimited data plans?”

Allot Communications’ John Priest said service delivery is becoming far more personalised and tailored than in the past, suggesting operators should have a think about how they allow users to access the information and the content they want.

“I think we need things like VAS for the customer to make them feel in control of what they need. It’s not just the volume, it’s about the content they want and delivering that to them. Users want to know what they’re consuming, so the SP has to know what they’re consuming and how they’re consuming it, so that you can be more proactive in the customer care and guarantee a higher quality of service.”

“But here are a couple of examples of opportunities for SPs to generate revenue aside from traditional voice and data pricing models. Sponsored data solutions, for example, sees the user get free data if they’re going to certain sites, which is a partnership between the SP and the content provider. Similarly to music streaming services, SPs can strike up a partnership with advertisers.”

US MVNO FreedomPop’s Mauricio Sastre suggested there are opportunities for partnerships between carriers, app developers and content providers, help users consume new applications for reduced costs, or for free.

“In an effort to get their app out there, the app developers are considering paying for the data that their app consumes, so there can be a subsidised way for users to consume their application.”

Of course, when debating content services over service provider networks, we naturally reach the inevitable pain point of net neutrality, Wireless 2020’s Sarkissian rounded of the discussion on such a note.

“When do we get into the net neutrality debate, the big elephant in the room? Do you decide to charge more for content from Netflix, or NFL Now, or WebRTC services?” he said. “In my opinion, any resource that’s finite, and has the potential to be fully consumed, should be left alone from regulation. Otherwise you jeopardise the investment into, and the competition of, all of these services.”


U.S. seeks trials to test transition to digital phone networks

31 Jan


U.S. wireless providers like AT&T Inc andVerizon Communications Inc on Thursday received a nod from regulators to test a transition of the telephone industry away from traditional analog networks to digital ones.

The Federal Communications Commission unanimously voted in favor of trials, in which telecommunications companies would test switching telephone services from existing circuit-switchtechnology to an alternative Internet protocol-based one to see how the change may affect consumers.

The experiments approved by the FCC would not test the new technology – it is already being used – and would not determine law and policy regulating it, FCC staff said. The trials would seek to establish, among other things, how consumers welcome the change and how new technology performs in emergency situations, including in remote locations.

“What we’re doing here is a big deal. This is an important moment,” FCC Chairman Tom Wheeler said. “We today invite service providers to propose voluntary experiments for all-IP networks.”

The move in part grants the application by AT&T to conduct IP transition tests as companies that offer landline phone services seek to ultimately replace their old copper wires with newer technology like fiber or wireless.

“We cannot continue requiring service providers to invest in both old networks and new networks forever,” Commissioner Ajit Pai, a Republican, said.

Some consumers, particularly in rural or hard-to-reach areas, have complained about poor connectivity of their IP-based services. Advocates have also expressed concerns about the impact of the transition on consumers with disabilities.

“I think we must be mindful of the impact this transition has on consumers — their needs, their expectations and their willingness to embrace network change,” said Commissioner Jessica Rosenworcel, a Democrat.

The trials will be voluntary, and regulators require that the experiments “cover areas with different population densities and demographics, different topologies, and/or different seasonal and meteorological conditions.” They also require that no consumers be left disconnected.


Evaluating ROI on Social Media for Telecom Service Providers

21 Nov

Which social media website to choose based on certain user specific criteria (Created using Gephi)

Which social media website should be chosen? What is the Return on Investment? (Picture Created using Gephi based on list of social media websites from wikipedia & calculating weighted mean)

Most telecom service providers use IT in key business processes like Marketing, Sales & Service. It would be worthwhile to integrate social media listening capabilities with the current systems so that the social media impact on the business can be measured. Below are a few metrics I could think of that can be used to understand if the social media strategy used is a successful one.

Customer Readiness for Social Media Influence – Marketing

Customer readiness to accept the influence of social media in their decisions to purchase products from the service providers can be understood with the following metrics. Based on the metrics given below, the social media campaign / channels can be decided upon.

  • During customer care / dealer / web portal interactions, are your customers ready to share their social media handles?

  • Which social media handle data are your customers more willing to give (like facebook, twitter, google+, linkedIn, etc.,) and our activity on the particular site.

  • What % of the entire customer base’s social media handles do you have?

  • What is the increase in the readiness of the customer to share their social media data?

  • How many customers interact with you on social media (like / favorite / comment / share)

 Branding – Marketing

Brand perception relies a lot on the products the service provider sells and hence it is important for the listening engine to understand what customers (prospective & current) think about:

  • the products released recently and the positives / negatives through sentiment analysis

  • the most popular products sold (according to the users)

  • the reasons associated with the segment-wise popularity of the product.

  • No. of times a product has been discussed on social media and its impact on sales.

  • Influence on branding by those who are not our customers or unidentified as customers.

 Campaign Management – Marketing / Sales

Nowadays we see a lot of campaigns being created exclusively for those on social media. It is important to keep track of

  • the number of campaigns that have been created based on the requests / interests identified through social media.

  • The number of campaigns created and the revenue out of these campaigns should give an idea of the returns on investment.

  • Map enterprise level data with the campaign data to understand the segment which likes the campaign.

  • How successful are the campaigns through Social Media? Revenue vs cost from these campaigns?

 Generating leads – Sales

While it is good to interact with customers and enhance brand awareness and influence, if the goal of the service provider is to actually find leads and convert them into business, the following are very important:

  • How many people have you converted into leads through Social Media?

  • the % of leads that could be contacted through Social media / outside of Social media.

  • How much % of these leads have been converted into actual customers by selling your products / subscriptions?

 Client specific goals & parameters – Service

The listening engine should take into consideration that each service provider / client base is unique and include the client specific goals into consideration before posting content across the social media. A few examples are given below:

  1. For example, if a telecom service provider wants to increase the usage of self-service portals thereby reducing the customer service requests through customer care channel, the company needs to devise logic to convert the goals to measurable ones & spread awareness through social media

  2. How much time can the service provider afford to spend on social media.

  3. If the customer base is multilingual, are their messages routed to the right support agents? Or is there an internal re-assignment?

Complaint Handling – Service

The number of issues / complaints identified gives a sense of the effectiveness of the listening engine. All identified issues / complaints could be saved as service requests / trouble tickets based on the nature of the issue. It is important to understand how many service requests raised are actually solved so that it can be compared to the traditional way of raising service requests. The Service Requests raised through the listening engine are initially not tagged to any of the accounts (remember “eventual consistency” in previous article?). The approximate account – subscription details to be tagged to this request is maintained in other fields and the customer care agent manually accepts or changes the account tagged. If the information cannot be tagged to any of the accounts, a personal message is sent to the user for the details and responded to their post on the same channel.

  • Number of service requests / trouble tickets have been identified through social media

  • % of service requests / trouble tickets solved

  • % of solved issues required a follow up customer care call VS how many were solved using information in social media itself

  • % of anonymous Service Requests that are later tagged to accounts & channels used as mentioned above

Churn reduction % – Service

Churn is the amount of customers moving from our network to other networks. Social media can help the service provider in understanding the ported out customers by listening to them on social media. The impact of social media on churn reduction can be found by looking at the following metrics:

  • The churn ratio among those people engaged on social media to people not engaged on social media. Higher the ratio, lower is the impact of social media influence on churn reduction

  • the number of Number Portability Requests due to social media. (This could happen due to competitors or due to influencers on social media who influence our customers to move out to another network)

  • Approx. number of customers where social media interactions were helpful to prevent churn beforehand

  • Positive / negative of feedback of customers who have ported our from a network on social media

 Competitor Analysis – Marketing

Most metrics defined here need to be monitored for our competitors as well. This gives a better perspective of the strengths and weaknesses of competitors, as perceived by the direct customers. As we monitor, we could scale up our strengths and also grow in areas considered popular by the end users. The perception & changing perception of our own strengths and weaknesses as perceived by people can be monitored on a regular basis and checked if we are moving in the right direction. The number of customers identified because we are monitoring our competitors can also be noted on a continuous basis.

One more important metric is the number of voices of those who popularize our competitors and their influence rate. These will help us understand the kind of people our customers listen to, in addition to us and our competitors.

Big Data is not about others success stories, but ours!



7 Nov
Using MPLS Protocols to distribute Network Virtualization Overlays across a Service Provider Cloud

Server Virtualization and the evolving network requirements has led to a recent rise in Network Virtualization Overlay (NVO) solutions. VLAN’s and MPLS is a current network virtualization method but has perceived shortfalls with scale and complexity, thus giving rise to new and innovative network virtualization (NV) technologies VXLAN and NVGRE.

Both VXLAN and NVGRE are layer 3, IP-based technologies that pre-pend an existing Layer 2 frame with a new IP header, providing layer 3 based tunneling capabilities of Layer 2 frames. Essentially, this means you can extend a Layer 2 domain across a Layer 3 boundary.

Scenario 1: Network Virtualization Overlay for Fully Virtualized Environments

In the diagram below, VM1, VM2, and VM3 are all part of the same Layer 2 VLAN, say VLAN 45, because the Layer 2 frames are encapsulated via the VXLAN (we would also use NVGRE, but we’ll focus solely on VXLAN) IP header.  Since VXLAN is a Layer 3 protocol, this effectively enables the Layer 2 domain to be distributed across the Layer 3 boundary.


The VTEP includes an important function, to consult it’s VM to VXLAN ID VTEP mapping table and inspect Layer 2 frames or Layer 3 packets and determine if a VXLAN header should be added or removed.  The VTEP does the encapsulation and de-encapsulation.  Notice also in the figure above, the VTEP function is housed in the Hypervisor for this scenario.

Scenario 2: Network Virtualization Overlay for Virtualized and Physical Environments

In this diagram, the scenario changes a bit. On the virtualized side, we still have the VTEP in the Hypervisor but on the right side, we have a physical server that’s been placed on its own unique subnet. The challenge comes when these two environments has a Layer 2 requirement to communicate via one another but are in different subnets in different parts of the network.  Here again, NVO’s and VXLAN can help.


The VTEP functionality, supporting the VM to VXLAN ID VTEP mapping table, can also be moved into the network switch. VLAN 45 can be associated with VM1 and VM2 in subnet A and also with the non-virtualized server in subnet C.  VXLAN provides the Layer 3 bridge between the physical and virtual environments. Note in the above diagram, the VTEP function can be in the switch.

Scenario 3: Network Virtualization Overlay across Service Provider networks

With the recent demand for new network virtualization strategies, another deployment model may include providing Network Virtualization across a Service Provider cloud.  For example, Layer 2 MPLS VPN’s or Layer 3 MPLS VRFs have been common service offerings for 10+ years so enterprises and customers are likely to continue to demand these service offerings, but now also using the perceived simpler and more scalable NVO technologies, including VXLAN and NVGRE.


In this environment, we have two network virtualization technologies that need to be aligned – VXLAN and MPLS. Please note that BGP and MPLS VPN’s are proven Service Provider Network Virtualization technologies, defined in RFCs 2547 and 4364.

The VXLAN VTEP identifies tunnel endpoints, provides the encapsulation and de-encapsulation of the VXLAN header, and provides the VM-to-VXLAN ID VTEP table mapping.

The MPLS PE router performs normal MPLS VPN functions, including using BGP for route distribution and VRF’s for unique routing tables and identified local attachments.  However, the local attachments here are the VXLAN IDs soa VRF associates with a specific VXLAN ID.  BGP is used to distribute the MPLS VRF labels among the MPLS PE routers while the MPLS P routers (MPLS Core) remain unchanged with MPLS label switching.

One interesting point is that VXLAN and NVGRE can support 16 million unique IDs due to the 24 bit wide header field.  The MPLS Label field is 20 bits wide, indicating there are approximately 1 million unique MPLS VPN ID’s.  This issue must be reconciled for any VXLAN to MPLS VPN mapping development.

Ultimately what is needed is an MPLS VPN to VXLAN mapping functionality that bridges the two Network Virtualization technology domains, aka L3VPN – VXLAN internetworking functionality.


A simple network

1 Feb
So i decided to play around with some Service Provider technologies and built myself a little test network with GNS3.

The design so far is really basic since i am just playing around with it.


The main network has 8 routers in it with routers 1,2,7 and 8 working as PE routers and 3,4,5 and 6 working as backbone. At the moment R9 and R10 are just simulating some client routers for a test.

Every router has its own loopback and for simplicity it is just based on the router number( for R1, for R2 and so on). As for the IS-IS configuration, atm i am only using a L2 network.

The configuration itself is not very hard, since basicly all you have to do is enable the process and give it a net address. Also since i am going to use this to mainly test some MPLS VPN features and some MPLS TE i set the metric style to wide. Here is an example from R1:

router isis 1
net 49.0001.0000.0000.0001.00
is-type level-2-only
metric-style wide
passive-interface Loopback100

Process enabled so i just needed to add this to the interface as well. By the way i am using normal ethernet interfaces because if i use a FE or GE interface i keep getting packet loss and malformed packets and apparently it is related to this version of Dynamips.

interface Ethernet2/0
ip address
ip router isis 1
duplex full
mpls ip

IS-IS is working correctly and we can see that the routing table is looking good
R1#sh ip route
Codes: L – local, C – connected, S – static, R – RIP, M – mobile, B – BGP
D – EIGRP, EX – EIGRP external, O – OSPF, IA – OSPF inter area
N1 – OSPF NSSA external type 1, N2 – OSPF NSSA external type 2
E1 – OSPF external type 1, E2 – OSPF external type 2
i – IS-IS, su – IS-IS summary, L1 – IS-IS level-1, L2 – IS-IS level-2
ia – IS-IS inter area, * – candidate default, U – per-user static route
o – ODR, P – periodic downloaded static route, + – replicated route

Gateway of last resort is not set is subnetted, 1 subnets
C is directly connected, Loopback100 is subnetted, 1 subnets
i L2 [115/20] via, Ethernet2/0 is subnetted, 1 subnets
i L2 [115/10] via, Ethernet2/0 is subnetted, 1 subnets
i L2 [115/20] via, Ethernet2/0 is subnetted, 1 subnets
i L2 [115/20] via, Ethernet2/0 is subnetted, 1 subnets
i L2 [115/30] via, Ethernet2/0 is subnetted, 1 subnets
i L2 [115/40] via, Ethernet2/0 is subnetted, 1 subnets
i L2 [115/40] via, Ethernet2/0 is variably subnetted, 9 subnets, 2 masks
C is directly connected, Ethernet2/0
L is directly connected, Ethernet2/0
i L2 [115/20] via, Ethernet2/0
i L2 [115/20] via, Ethernet2/0
i L2 [115/20] via, Ethernet2/0
i L2 [115/30] via, Ethernet2/0
i L2 [115/30] via, Ethernet2/0
i L2 [115/40] via, Ethernet2/0
i L2 [115/40] via, Ethernet2/0

So lets try to reach the other side of the network

Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to, timeout is 2 seconds:
Success rate is 100 percent (5/5), round-trip min/avg/max = 72/103/144 ms

And we have connectivity. But since we have mpls ip enabled on the interfaces are we really using mpls?

R1#sh mpls ldp bin
lib entry:, rev 2
local binding: label: imp-null
remote binding: lsr:, label: 16
remote binding: lsr:, label: 23
lib entry:, rev 8
local binding: label: 17
remote binding: lsr:, label: 17
remote binding: lsr:, label: 24
lib entry:, rev 6
local binding: label: 16
remote binding: lsr:, label: imp-null
remote binding: lsr:, label: 22
lib entry:, rev 16
local binding: label: 21
remote binding: lsr:, label: 18
remote binding: lsr:, label: 20
lib entry:, rev 20
local binding: label: 23
remote binding: lsr:, label: 20
remote binding: lsr:, label: 21
lib entry:, rev 24
local binding: label: 25
remote binding: lsr:, label: 22
remote binding: lsr:, label: 16
lib entry:, rev 30
local binding: label: 28
remote binding: lsr:, label: 25
remote binding: lsr:, label: imp-null
lib entry:, rev 32
local binding: label: 29
remote binding: lsr:, label: 26
remote binding: lsr:, label: 29
lib entry:, rev 4
local binding: label: imp-null
remote binding: lsr:, label: imp-null
remote binding: lsr:, label: 27
lib entry:, rev 10
local binding: label: 18
remote binding: lsr:, label: imp-null
remote binding: lsr:, label: 28
lib entry:, rev 12
local binding: label: 19
remote binding: lsr:, label: imp-null
remote binding: lsr:, label: 25
lib entry:, rev 14
local binding: label: 20
remote binding: lsr:, label: imp-null
remote binding: lsr:, label: 26
lib entry:, rev 18
local binding: label: 22
remote binding: lsr:, label: 19
remote binding: lsr:, label: 18
lib entry:, rev 22
local binding: label: 24
remote binding: lsr:, label: 21
remote binding: lsr:, label: 19
lib entry:, rev 26
local binding: label: 26
remote binding: lsr:, label: 23
remote binding: lsr:, label: imp-null
lib entry:, rev 28
local binding: label: 27
remote binding: lsr:, label: 24
remote binding: lsr:, label: 17

Well we have lables so it should be working….

R1# trace

Type escape sequence to abort.
Tracing the route to

1 [MPLS: Label 26 Exp 0] 72 msec 128 msec 100 msec
2 [MPLS: Label 28 Exp 0] 104 msec 100 msec 76 msec
3 [MPLS: Label 25 Exp 0] 64 msec 40 msec 104 msec
4 92 msec * 104 msec

This was just to build  little test network. From here i will be working with some BGP and a few VRFs.


The Internet Monopoly

13 Nov

Add your thoughts here… (optional)

Where and how to Invest in Telecom Growth

5 Nov

Investors who have had their hard earned money invested in telecom have had a rough few years.  What once was a lucrative growth story has turned into a high-risk adventure that for some reason still has people calling the remaining skeletons value plays.  Alcatel-Lucent?….No thanks.  Ericsson?…..Ugh.  Nokia Siemens?….I don’t think so.   Motorola?….right.  Nortel, Marconi?….don’t exist anymore.  Cisco?….what have you done for me lately?

A boatload of factors has contributed to the demise of these former high flyers.   Lack of focus by management trying to be everything to everyone everywhere, macro-economic issues causing drastic reduction of Communications Service Providers (CSPs) CAPEX budgets, fierce competition by the entry of Huawei and ZTE driving prices and margins down, using what many suspect are unfair practices and government support.  Valid reasons, but not the complete story.  There is an often overlooked issue that I think should be the biggest concern:

Although they will never publicly admit to it, the CSPs simply do not consider the networks as a differentiator anymore.  And it so happens that despite all the other solutions and services that Alcatel-Lucent, Ercisson, Nokia Siemens, and even Huawei and ZTE have to offer, networks remain their core business.  Unfortunately for them, CSPs’ networks are now considered by their owners to be commoditized pipes that do not contain the intelligence that is required to create a competitive advantage.   That is the real reason why margins are so compressed.  CSPs are just not prepared to pay the premiums they once did.

The problem is that the above-mentioned corporations are, despite all the reorganizations, still built as if their network equipment solutions sell at premium pricing.  They don’t anymore.  In reality, network infrastructure solutions have become a commodity business.  I am staying away until I see them really re-invent themselves, and that is going to take years.

So what else is there if you are a growth investor and look for telecom opportunities?  You can have a look at some of the smaller vendors such as Infinera.  Lots of promise for a long time, but will they ever deliver against that promise?  Maybe worth a small bet?….maybe.

ATT, Verizon?  You have to look elsewhere for growth, unless they decide to seriously expand internationally, which would be out of character and unlikely to be in their best interest.

If you want to find growth in telecom, you have to look at Latin America and China.  I prefer Latin America as there is less of an issue with government involvement and quite frankly, Latin American telecom is doing much better,  A good indicator is troubled Alcatel-Lucent Q3 earnings report in which they highlighted the one bright spot: double-digit revenue growth in the Caribbean and Latin America.  Why is that?

Most countries in Latin America only have about 30% high-speed internet penetration rate mostly limited to urban areas.  This penetration rate is not due to lack of demand, but rather due to incumbent operators, so far, not seeing the business case in expanding their networks to the non-urban population

Something is changing though.  Improvements and expansions in high-speed internet infrastructure are directly related to economic progress.  Governments realize that these investments are necessary to continue to advance their development.  As a result, across the region they are now heavily promoting and investing in national broadband and fiber network deployments so that their people can access the internet.

Due to this lack of investment, in many Latin American countries, the telecom sector is heavily fragmented.  It is very common to see incumbent operators focusing on the densely populated urban areas and neglecting the rest of the country.  You will find within, but especially outside of, the big cities, tens or hundreds of small local cable operators. For example, Ecuador has over 200 and Paraguay more than a 100 cable TV operators.  But what these local operators offer are video services, not high speed internet.  So these cable TV operators have thousands of customers, existing revenue, they truly understand their local markets and they know they can sell broadband services into the majority of their customer base, if they have the budget.   This is similar to the environment that existed 20 years ago right here in the USA.

Here in the USA, twenty or so years ago, ComCast built a powerhouse by consolidating regional players under one umbrella.  I think that is exactly what will happen over the next decade in many of the Latin American countries.

Two companies seem to be well positioned for this wave of consolidation:

1.  America Movil (NYSE: AMX)

This is the Mexican telecom top dog in Latin America.  Owned by the world’s richest man – Carlos Slim, who also sits on the United Nations broadband commission that is pushing the agenda of enabling 60% of the population of all third world countries access to high speed internet at pricing less than 5% of GDP per capita.  AMX Operates in 18 markets across Latin America and the Caribbean where in most cases they have the #1 or the #2 market share. They are expanding their reach worldwide by acquiring larger positions in European operators such as KPN (The Netherlands) and Telekom Austria.

More importantly, with the increased push by countries to execute on their national broadband plans combined with the existing fragmentation in many of the markets AMX operates in, they are ideally positioned to lead the next inevitable wave of consolidation.  AMX has the financial power to instigate the consolidation.  And if governments are prepared to subsidize some of the investment to realize their national broadband plans AMX can create a profitable business case.  Combined with an accelerated growth of the region’s middle class who will have more money to spend on both basic and more sophisticated telecom services, AMX is poised for growth over the next 10 years.

2. Liberty Global (Nasdaq: LBTY)

Another company that may even be more likely to jump in the middle of the Latin American consolidation wave is Liberty Global, one of the leading international cable operators in the world with their headquarters in Colorado.  I really like LBTY.  They are opportunistic, patient, and decisive.  Always looking for expansions.  11 of their 13 operations are in Europe, with the remaining two in Puerto Rico and Chile.   They expanded in Europe by taking advantage of their own industry consolidation wave.  LBTY has been there before.  They know how to do this, which is an advantage they have over AMX, as AMX’s background is a legacy of being the incumbent, not the challenger.

With the ongoing crisis in Europe, I think their investment dollars will for the majority be redirected from Europe to Latin America.  They have had a few years of familiarizing themselves with doing business in the region, a necessary step that is often overlooked by many North American and European companies aiming to do business in Latin America.   The acquisition of VTR, the leading operator in Chile, in my opinion will prove to be a great launching pad to expand their presence in the region.  I cannot see them buying into just two operations in a high growth region without the intention to replicate their successful European business model.

Source: By Hans el Fasid – November 4, 2012 | Hans is a member of The Motley Fool Blog Network — entries represent the personal opinions of our bloggers and are not formally edited. I do not own shares in any of the companies mentioned in this entry.

%d bloggers like this: