What the heck is MEC?

October 13, 2015

cowboyFor those who have been following our posts for the last few years, we have been discussing three emergent trends:

  1. The blurring of the lines between mobile operator and enterprise infrastructure.
  2. The reversal of common IT platform services spend from 80% capex/20% opex to 20% capex/80% opex.
  3. The innovation opportunity created by mobility and the immense untapped capability of mobile operating systems to transform business.

Each trend is discussed in the context of computing needs in the cloud services offerings and, in the near future, edge processing inside the enterprise. The use of edge processing has been slow to emerge due to a lack of standards and the need to accumulate enough interest to catalyze the mobile operator market.

In the last year, the European Telecommunications Standards Institute “ETSI” kicked off an initiative to more fully develop Mobile Edge Computing “MEC” (pronounced MECk) and create a common technical landscape. This will allow the Independent Software Developer “ISV” community to more easily develop packages that are portable to any RAN vendor’s MEC hardware platform. You can learn more here. This is exciting to us as it’s a further proof point that the journey that SpiderCloud has been on to create an awesome customer experience along with critical services has been embraced by mobile operators and their key suppliers.

SpiderCloud spoke at the first MEC Congress, in London on 29-30 Sept 2015, about our early prototypes and experiences with our E-RAN in learning about the market and technology potential in the past five years. While many of the present considerations around MEC are focused on foundations for 5G services, SpiderCloud has been showing enterprise-focused services targeted at increasing revenue directly, or as feeder service that must be performed locally.

A few examples could be helpful to illustrate them:

  • Increasing revenue: Unified Communications tying the native interfaces on a mobile device to the enterprise’s infrastructure have drawn interest from enterprises that wish to integrate and simplify the lives of their business customers who are highly mobile. This reduces the need to build “enterprise OTT”, and simplifies day-to-day usage.
  • Feeder Service: Location Based Services raw events are capable of generating a very large stream and can impact backhaul if it’s sent in raw form to a cloud service. We collaborated withVodafone and HP to build an on-board x86 Services Blade and App for the E-RAN’s Services Node. The process of events stream into UE locations could easily be handed to an external application using the context of UE locations to make better decisions within an equipped facility. The edge service was necessary due to both time critical events, and the sheer volume.

The E-RAN’s ability to conduct many of these early prototypes is directly related to the Services Node aggregating the cloud of radios in the building. All mobility events are “in front” of the x86 Services Blade such that mobility event handling and context switching don’t need to be incorporated into the macro-network. Further, the cost of x86 processing capacity is amortized across all the Radio Nodes in the building. This makes the business case work. If every small cell needed an onboard processor, edge computing could easily be cost prohibitive for most applications.

And, finally, 451 Research’s Ken Rehbehn moderated a webinar on MEC development with Intel’s Caroline Chan, McAfee’s Dan Frey, and SpiderCloud on 8 Oct that is available for replay here. It explored MEC concepts, SpiderCloud’s early edge cloud experiences, a deep dive into operating the McAfee NG-FW as an enterprise service, and a number of audience questions about MEC.

So, we welcome ETSI MEC as yet another proof point that the trail we are blazing with our early mobile operators in both R&D and deployment is going in the correct direction. As MEC applications hosting and use cases develop, we suggest that Product Managers in mobile operators pay attention to the indoor applications market as it’s very unique and potentially far more profitable, relative to MEC in the macro-cellular.

– Art King, SpiderCloud Wireless, Director of Enterprise Services & Technologies

Twitter: @EMobilityInside
Visit our Enterprise IT site @ http://SpiderCloud.com/EInsider


Kista plays catch-up… again!

September 22, 2015

shoosewisely_0So, efforts to target the enterprise IT buyer by the mobile infrastructure industry received another strong validation last week. If you recall, SpiderCloud’s strategic collaboration with Cisco to deliver compelling offerings to enterprise customers was unveiled at MWC15. Well, it looks like Ericsson has finally realized that their “carrier-grade Wi-Fi” is not good enough for enterprise, and HP has realized that they cannot turn a blind-eye to licensed spectrum. So, here they are with a new press release.

17 Sep 2015 – Ericsson Press Release

Highlights:

  • Bilateral reseller agreement opens up enterprise market opportunities beyond Ericsson’s carrier-grade Wi-Fi and provides go-to-market opportunities for Ericsson’s industry-leading small cells through HP’s global enterprise channels
  • Leverages HP’s acquisition of Aruba Networks through integration of HP’s enterprise WLAN technology into Ericsson RBS 6402 picocell

There are a number of apparent gaps in the technology collaboration that appear problematic and make us ask questions, such as:

  • It only covers the RBS 6402, which is a picocell product targeted at the SMB market (up to 5,000 m2 building). The medium-large building “Radio Dot” platform is visibly missing from the announcement. Is HP-Aruba Wi-Fi in-scope for the Dot, or does a parallel infrastructure need to be put in place?
  • Where is support for enterprises who have HP-Aruba systems in-place now? We don’t see it. By contrast, current generation Cisco AP’s sport a modular port where our jointly developed cellular Clip-On module can be simply plugged into them. This provides for an implementation that leverages the pre-existing enterprise transport infrastructure, to quickly bring critical mobile service to the Cisco enabled buildings.
  • What are the RF deployment tradeoffs for implementation to stay within the standards based PoE+ power budget? From a product perspective, will there be multiple versions of the RBS 6402 that include the different mixes of 3G, LTE and Wi-Fi radios, or only one version? And, how many radios can be active?

What’s our take on the commercial effects?

The reseller agreement validates that mobile wireless infrastructure vendors have to find a way to offer their products directly to Enterprise. We have seen this trend emerging over the last three years as mobile has become business critical, and energy efficient (metal coated) window glass rollouts are accelerating.

While this transition was started with the emergence of scalable small cells that are “IT friendly” and achieve a Wi-Fi price point, the completion of transition will be stable, operator models that facilitate “frictionless adoption” by enterprises. This change in thinking is at the vanguard of many operators whose customers are demanding better service. We documented some of the building blocks to the future in our post Enterprise Small Cell Deployment Insights as a start point for commercialization thinking about the transition.

Finally, the technology exchange involved in the agreement is not clear on the roles of each of the Wi-Fi divisions. The need to seek an external source for 802.11ac technology for the RBS 6402 raises red flags. On the RBS 6402, there are additional new questions beyond our Small Cell “Super Bass-O-Matic’76”? post that unpacked the technology realities of the first RBS announcement.

As always, nothing is simple when planning for the complex landscape we live in globally.

– Art King, SpiderCloud Wireless, Director of Enterprise Services & Technologies

Twitter: @EMobilityInside
Visit our Enterprise IT site @ http://SpiderCloud.com/EInsider