The New Mobile Network
, Regular Visitor
The New Mobile Network
Six Network Predictions for 2015
Jan 13, 2015

The first 14 years of this century have seen massive transformation in the networking space, with the creation of products, applications and services exploding onto the market. The first iPhone launched putting more compute power in our pockets than the computer used on the Apollo missions. Facebook was born, connecting more than 1.3 billion people worldwide. And Google indexed the online world, leading to trillions of searches each year. All of this and more has of course led to service providers needing to build high-performance, massively scalable infrastructure simply to keep up with user demand.

pillar_network.jpg.jpeg

 

But the next five years promise to be as disruptive for network providers. Virtualization of the network reflects a shift towards an agile, dynamic IP platform for cost-effectively meeting demand while responding to, and profiting from, customer requirements. Below are a few of the trends Juniper sees taking hold in 2015 – ushered in through virtualization throughout the network.

 

IP Video Streaming Disrupts Cable Providers

With the exception of the switch to digital several years ago, the cable industry has largely remained the same for several years. 2015 will be the year that this changes. With the advent and rise of streaming content online, we’re already seeing signs that subscribers are demanding a whole new range of IP services and really high speed bandwidth from their cable providers as they shift towards different types of content providers such as Netflix. According this year’s U.S. Digital Video Benchmark from Adobe, viewership of streaming content online has risen nearly 400 percent since last year. And with HBO recently announcing it would offer its streaming app as a standalone service, things seem to be headed that way. To stay competitive, we anticipate 2015 being the year that cable companies make their bets on virtualized networks to cope with the expected increase in IP traffic. Although the transition will take several years, cable providers will use 2015 to assess the options and begin to put together RFPs to find vendors that match their vision.

 

High Frequency Trading is Waning, Here Comes Hyper-Contextual Trading

High Frequency Trading is declining, which is down from 70 percent of equity volume to less than 50 percent. Hyper-contextual trading (HCT) is the next disruption in the market. HCT is the real-time assimilation of both classic news feeds (Bloomberg, Thomson-Reuters, AP, CNN) and social media feeds (Twitter, Facebook, LinkedIn, Blogs, etc.) in order to exploit market news for competitive trading advantages. It will all be driven by analytics to ingest, process, and derive insights at speed to exploit market discontinuities. HCT is about distributed computing and performance; latency is table stakes and no longer a differentiator. Firms and their IT environment will need to pre-digest hundreds of information feeds in real time; this will require highly specialized programming and network equipment.

 

Product co-creation comes to networking: new ideas require new thinking

For as long as networks have been around, the development and sale of networking products has been more or less a one-way street: equipment vendors develop products – routers, switches, etc. – and sell them to enterprises and service providers. Enterprises and service providers bought them because they had to. In 2015 this will change. More large companies and service providers will become more vocal about the features and prices of the products they buy for their networks. For example, AT&T’s Domain 2.0 initiative and InterCloud Systems (TTM), are partnerships with the goal of co-creation. Expect to see more products in 2015 co-created between partners, customers and network equipment makers, as customers wield more influence in the development of the modern network.  

 

Big data and networks: opportunity or catastrophe?

As the tentacles (read: devices) of the so-called Internet of Things continues reaching into new parts of our lives, there will simply be a lot more data generated. For example, even the simplest connection between phone and home security systems will generate data that needs to be stored somewhere. What this will bring in 2015 is threefold: a newfound need to analyze this data, the required network infrastructure to make sense of it all and security technologies to secure it. Demand for data scientists will reach a fever pitch as service providers and enterprises rethink the way they build networks to handle the onslaught of data.

 

Ready for 5G?

When wireless carriers moved from 3G to 4G offerings, it was arguably the first time customers could viably expect to stream content on the go. More and more providers and technology companies are exploring 5G, but are the networks ready for it? And moreover, what will 5G actually enable beyond what we have now? Study after study has shown that increased mobile speeds means people simply consume more. As unlimited data plans dwindle and more connected devices come online (think a toaster that texts you when it’s ready), will people be willing to foot the bill for the next-gen speeds of 5G? These are the questions service providers will need to ponder in the coming months.

 

The Rise of HD... Voice Calls?

As 4K-quality video content poses challenges for network capacity, there’s another (perhaps less sexy) HD technology brewing: voice calls. Voice over LTE (VoLTE) is starting to crop up in feature lists for new smartphones, but in 2015 we anticipate consumers will start to finally notice a difference in voice quality between regular phone calls and VoLTE calls. No more, “Wait, say that again.” Concurrently, we’ll have to wait and see how such a service will impact both network capacity and security.

 

You can find the full Top Network Predictions for 2015 graphic on Slideshare. Find more of Juniper Networks’ 2015 predictions by following @JuniperNetworks.

Jan 26, 2015

Steve,
firstly I would like to congratulate you on an insightful blog post.
The section which grabbed my own attention was your paragraph entitled: "High Frequency Trading is Waning, Here Comes Hyper-Contextual Trading". Whilst there is much I agree with here, let me offer a slightly different perspective. There remain a range of differing trading strategies. Trading strategies which are based on signals with a fast rate of decay require a significant investment in order for the investor to remain competitive. For those who cannot afford to be at the sharp end of this latency arms race, there has been a more recent trend in strategies which are based less around fractional differences in pricing of a security at a given point in time and more on big data analytics. This is not to say that ultra low latency trading at high frequencies ceases to be relevant. There remains a continued need for infrastructure to support the transmission of data with minimal latencies. Innovation in that area needs to continue to support a demand in a more tightly regulated trading ecosystem.
What I believe has changed is the introduction of alternate trading strategies, offering greater choice for investors. The widening of choice in available trading strategies driven by innovation in big data and analytics is likely to be a contributing factor for the proportionate reduction in trading volumes driven by HFT.
Thomson Reuters explain it quite nicely in their blog post entitled: "latency - dead, or just misunderstood" (http://blog.financial.thomsonreuters.com/latency-dead-just-misunderstood/).
So I would counter that the need for ultra low latency infrastructure remains very relevant even if it's less prevalent.

Perry

Feedback