Tuesday, August 26, 2014

SDN / NFV part V: flexibility or performance?

Early on in my investigations of how SDN and NFV are being implemented in mobile networks, I have found that performance remains one of the largest stumbling blocks the industry has to overcome if we want to transition to next generations networks.

Specifically, many vendors recognize behind closed doors that a virtualized environment today has many performance challenges. It explains probably why so many of the PoCs feature chipset vendors as a participant. A silicon vendor as a main proponent of virtualization is logical, as the industry seeks to transition from purpose-built proprietary hardware to open COTS platforms. It does not fully explain though the heavy involvement of the chipset vendors in these PoCs. Surely, if the technology is interoperable and open, chipset vendor integration would not be necessary?

Linux limitations

Linux as an operating system has been developed originally for single core systems. As multi-core and multithreaded architectures made their appearance, the operating system has shown great limitations in managing particularly demanding data planes applications. When one looks at virtualized network function, one has to contend with the host OS and the guest OS. 
In both cases, a major loss of performance is observed at enter and exits of a VM and the OS. These software interrupt are necessary to pull the packets from the data plane to the application layer so that they can be processed. The cost of software interrupt for kernel Linux access ends up being prohibitive and create bottlenecks and race conditions as the traffic increases and more thread are involved. Specifically, every time the application needs to access the Linux kernel, it must pause the VM, save its context and stall the application to access the kernel. For instance, a base station can have over 100k software interrupt per second.

Intel DPDK and SR-IOV

Intel Data Plane Development Kit (DPDK) developed by 6Wind is used for I/O and packet forwarding functions. A “fast path” is created between the VM and the virtual network interface card (NIC) that improves the data path processing performance. This implementation allows to effectively bypass the guest hypervisor and to provide fast processing of packets between the VM and the host.
At the host level, Single Root I/O Virtualization (SR-IOV) is also used in conjunction with DPDK to provide NIC-to-VM connectivity, bypassing the Linux host and improving packet forwarding performance. The trade off there is that each VM on SR-IOV requires to be tied to a physical network card.

Performance or Flexibility?

The implementation of DPDK and SR-IOV has a cost. While performance for VNFs implementing both techniques show results close to physical appliance, the trade off is flexibility. In implementing these mechanisms, the VMs are effectively bound to the physical hardware resource they depend on. A perfect configuration and complete identical replication of every element, at the software and physical level is necessary for migration and scaling out. While Intel is working on a virtual DPDK integrated in the hypervisor, implementation of SDN / NFV in wireless networks for data plane hungry network functions will force vendors and networks in the short to medium term to choose between performance or flexibility.

More content available here.

Thursday, August 14, 2014

Impact of Iliad's purchase of T-Mobile

Last month, I was musing about a world where wireless voice and messaging ARPU would be 0$.

While many are discovering the bid by Iliad to take over T-Mobile US, I have been following the company for nearly two years in its successful market introduction in France.

I thought it would be interesting to have a look at what has been the impact to date of Iliad on the French mobile market as a reference point.

Iliad launched a mobile service in France in January 2012 under the brand free Mobile. The offer was simple and a perfect disruption to the highly-regulated market. Building on their broadband payTV set top box offer, Iliad secured 3G and 4G licenses from the french regulator and started offering for free mobile services to their payTV customers.

The offering

Shortly thereafter, the company launched a disruptive offer: 19.99 euros per month for unlimited national (all) and international (fixed line) voice calls , unlimited text and picture messages in europe, 3GB of data in 3G and 20GB in 4G.
To accelerate their customer acquisition for the moderate users, the group launched in 2013 a no contract 2 euro per month deal for 2 hours national and international voice calls, unlimited text and picture messages in europe, and 50MB of data.

The results

This disruption was a commercial and popular success for the company but a disaster for the incumbents. 
Dominated by Orange, followed by SFR and Bouygues, the market was deeply disturbed by this introduction.

In two years, here is the impact of Iliad on the french market:

  • Iliad as of March 2014 counted 8.6 million subscribers representing 13% market share and generating 1.2 billion euros of revenue.
  • Iliad in 2014 covers 50% of the french population in 4G and 75% in 3G.
  • Iliad's network quality has been rated worse of all French operators but the company ranks first in customer satisfaction.
  • Average turnover by mobile operators decreased by 11% in 2012 and 13% in 2013 and net income by 20%
  • Average Revenue Per User (ARPU) has decreased by 22% 
  • Collectively, operators increased their CAPEX investments to 7.3 billion euros in 2013 (excluding licenses).
  • Collective free cash flow has decreased by 40%


Of course, it would be difficult to draw a direct parallel between the successful introduction strategy of Iliad in France and what would happen if they purchased T-Mobile in the US. Nonetheless, Iliad is probably the company that has acquired the most customers in the shortest amount of time of all wireless operators globally, while focusing on what matters most: customer satisfaction. Of course a disruptive pricing strategy was the main vehicle for introduction, but ease of use, with no-contract offers, unlocked phones packaged or not with subscription were a large part of the company success as well. 

The lessons to draw here are that network operators need to prepare for a world where ARPU can drastically reduce while structural investments increase. Flexibility, elasticity but more importantly a customer centric approach will make the difference.
The themes are further addressed and analysed in my latest report to be released in September: SDN / NFV in wireless networks.

Tuesday, August 12, 2014

Why my kickstarter project is going to fail

I have recently launched a kickstarter campaign for my next report "SDN - NFV in wireless networks". I think I might be among the first/only analysts that are using Kickstarter for this kind of project.
I provide highly specialized labor and intellectual property-intensive reports where I strive to paint an independant picture of an ecosystem, drawing out its dependencies, challenging its operating assumptions and understanding its potential and weaknesses.
Kickstarter is well suited for projects that have mass appeal and capture people's' imagination. While I find SDN and NFV in wireless networks fascinating, I don't think that this interest really qualifies as mass market appeal. My children keep telling me that I will fail making NFV in wireless network attractive if I can't draw parallels with NFV and the TM forum's catalyst projects...
More seriously, this is a B2B enterprise, targeted at the relatively small niche of companies commercializing or investigating SDN and NFV in wireless networks. It is not, generally speaking, an audience prone to serendipitously support projects. While the report and other associated deliverables I develop provide tremendous value for these organizations, this is simply not the way they source these reports and it is unlikely that many will support the project by pledging.
Second, Kickstarter is based on the concept of incremental rewards for project backers. This project might be mostly monolithic with a report as its central deliverable and the comparatively high price compared to other kickstarter projects might be detrimental to its target. I deliberated long and hard on the different backing levels, their discounted price vs retail, their value, the limit I want to impose on them...
I am very grateful for the signs of support and pledge received to date, but all in all, this vehicle might not be the best for the financing or launch of this type of reports. Good thing I am not relying on it. This is more an experiment in social engineering than an actual attempt at financing the project. I am fortunate that my previous reports have been doing very well and they generated tremendous support and recommendations.
The reality is that I am writing and delivering this report and associated deliverables even counting on the kickstarter campaign to fail. The report is already largely pre-sold and financed. I will be chairing and keynoting the SDN & NFV USA show and the SDN - NFV in mobile track at the SDN & OpenFlow world congress in October where I will officially launch the report and present some preliminary findings.
Why did I do it then? Well, it is good every now and then to do something new. Something no one has done before. Failure is not a curse, it is sometimes desirable, if you plan to learn from it, it leads to great success.