Much has been made about the revival of the telecom industry and of the latest and greatest technologies that are going to come back into style. To be successful in today's market, however, products need to be almost immediately able to produce definable operational efficiencies, near-term compensating revenues, or some sort of strategic advantage that can be leveraged in service offerings. This can be challenging for equipment vendors and service providers.
There are a lot of curve balls being thrown in the market, and this will impact the viability of many of the present business strategies by telcos worldwide. Here are 10 on the leading edge of having a major impact on opex, capex and strategic decision-making in the near term.
Wi-Fi hot spots have been the "next big thing" for a couple of years now, but no hot spot operator has deployed (or even proposed deploying) an array of hot spots with sufficient coverage to provide ubiquitous coverage within a metro area. The cost of deploying Wi-Fi access points with such coverage has thus far been prohibitive -- even as Wi-Fi prices plummet -- mainly due to the cost of providing backhaul to the hundreds or even thousands of access points needed to provide ubiquitous coverage in a larger city.
Metrospots (or hot zones) can use one of two emerging technologies to provide citywide coverage at a lower cost. The first option is to deploy a portable wireless system that can provide citywide coverage while requiring only a handful of towers -- and associated backhaul facilities. The other option is to leverage meshed Wi-Fi systems that offer standards-based Wi-Fi connectivity to end users while providing self-discovering and self-healing meshed connectivity back to a smaller number of wired backhaul locations.
Either of these approaches may offer faster than 2.5 or 3G mobile systems data connectivity within a metro at a significantly lower price. In particular, the meshed Wi-Fi approach, combined with Wi-Fi enabled mobile phones, can provide a compelling alternative to next-gen cellular data system deployments.
Managed Home Networks
We've moved on from the time where experts were prognosticating the "network-enablement" of every consumer device imaginable into the age where it's actually being done. And it's not just Internet refrigerators, but rather a lot of things that are a lot more mundane, like home entertainment systems.
More network-enabled devices means more IP addresses and more use of DHCP/NAT in the home. It means minimum levels of required bandwidth in the access loops and in the home network. It means showing transparency while at the same time allowing much broader flexibility to handle different types of devices, access rights, etc.
Part and parcel of having all these items in the home is the requirement to manage access at a physical layer at the least, and to be able to apply certain QoS levels to these connected devices. Without this, home network performance will be lacking for server and high-bandwidth applications.
In the past, the home gateway has largely stopped at the point of demarcation, since many wireless gateway devices are part of, or collocated with, the broadband access equipment where it enters the house. Newer technologies, mostly geared toward backward compatibility with existing phone and electrical wires or toward wireless solutions, have entered the home, but still a node-to-node ability to map and graph home network performance has been lacking.
But we're excited about a new level of functionality that is focused on extending the demarc to each room, through advanced home networking technologies. Such systems, exemplified by SerCoNet's NetHome system, enables you to have IP addresses and management at the room level, while offering a flexible, modular approach to enabling analog voice, Ethernet, FireWire, USB, 802.11a/b/g, VoIP, and other interfaces at the same phone plate that you offer regular phone service today. Such a wiring scheme opens up the potential for managed home network services within the home, and therefore a new revenue opportunity for service providers.
Wireless Handsets/Home Wireless Backbone
We've all heard about the falling costs for wireless devices, and how they are spreading everywhere. But one of the not-so-talked-about items is the emergence of a home wireless backbone. As the price of access points plummets, and the advancements in mesh networking in the home proceed to offer the ability for multiple access points in the house to work with one another, you start to have the establishment of a home wireless backbone. Once it's in place, devices and services start to emerge to take advantage of that environment.
A good example is cordless phones: Most people hate them because of their poor performance and interference issues. Moreover, as home wireless LANs emerge, there is a new interference mechanism for the cordless phone, since they may operate over similar frequencies.
Because of this, we expect cordless phones will quickly migrate to the 802.11a/b/g backbone, improving quality, improving reach, and providing more capabilities to the phone itself. We also expect a large take-up rate in 802.11/cellular multimode phones that will allow someone to log onto their home network when in range of their wireless backbone. This could well converge with the above movement away from traditional cordless phones. (Does the Linksys/Cisco combo make sense now?) And with the wireless backbone in place, more and more devices will hop on.
Does this mean you won't need a wired infrastructure in your home at all? We think you need both. We've done a lot of testing of applications on wired and wireless networks in the home, and when you start stressing these networks with streaming applications, video, and server access, you run into quality issues. Add a few teenagers and one gamer to the mix, and you've got a solid requirement for a 100Mbps home backbone pretty fast, and that's beyond the range of today's shipping wireless solutions. Wired and wireless should be symbiotic in the home, not competitive.
Windows XP Media Centre Edition
Within the home, convergence has moved from promise to reality. The advent of sub-US$1,000 PCs running Windows XP Media Centre Edition has brought together broadband, personal computing and home entertainment into a single consumer-friendly box. With a Media Centre PC, a consumer can watch DVDs; listen to CDs, MP3s and WMA music files -- including files obtained online -- view and print digital photographs, and take advantage of TiVo-like TV functionality -- all through a remote controlled interface on the TV.
The current Media Centre platform is weak in two areas: online content and support of home networking. Both of these areas are near the top of Microsoft's "wish list" for future functionality, and improvements here will provide a great opportunity for online service providers (and the broadband providers that enable their services) to begin to offer online entertainment services that move beyond the PC and into the rest of the house.
For consumers who don't want to invest in a new PC, similar functionality is being rolled out in set top boxes powered by software like Diego's Moxi (recently deployed to 100,000 set top boxes in Charter's network).
IP Carrier Services
Have carriers finally embraced the need for IP services in their networks? All have started down this path in various stages, but we believe the tidal wave is about to hit. The question is why is this happening now and what is happening.
The "why" is the easy part: competition and margin. As cable companies continue to expand their offerings, carriers must respond to keep their coveted voice customers buying new high margin data and video services. The "what" can take many shapes and sizes. Services such as VoIP, video and audio streaming, on-demand services to include audio, video, gaming and software applications, and the ability to control the user experience of each of the application streams sent to them.
Perhaps the biggest "new" news is the DSL Forum's WT-081 (about to become TR-059) effort which is the first major step in outlining how the carriers will utilise per-application flows to enable a host of new IP service offers to end users that will hopefully increase demand and margin. This set of technical requirements was crafted by major telcos for the expressed purpose of opening up their nets while staving off new capex investment to do so. It's a major move forward for IP services, and something to watch.
Lately, it's been popular for the press to bash active and passive optical network solutions. After all, they cost thousands of dollars per endpoint, require running fibre through neighbourhoods, and mandate changes in the way telcos even approach network engineering.
But the fundamental network necessities are driving the need for higher bandwidths at the network edge of the network, and fibre is the long-term solution everyone agrees is the right way to go, so the question is when to start, right? Well, sort of. The fact is, the telcos already started major telco deployments of fibre in the loop in earnest back in the 1980s when running fibre throughout the metro areas. And throughout the 1990s and into the 2000s, we saw a lot of the new neighborhood builds shift almost totally to fibre, at least in the major telco service areas.
Now, we're seeing pressures to expand beyond these specific business cases into the more mass-market business case for fibre to the premises -- a more general approach to pushing fibre to all sorts of residential and business establishments located in more suburban and other areas.
The triumvirate of BellSouth, SBC and Verizon and their RFP for PON is merely one milestone -- albeit a big one -- in the path toward pushing fibre closer to customers. BellSouth is already deploying more than 200,000 lines a year of fibre in its territories, and expects to jumpstart this to higher levels under this RFP. Indeed, we're at the neck of the hockey stick in terms of growth in optical access technologies. It will ramp quickly.
Storage prices continue to plummet, and this is going to mean a lot of changes in business cases for broadband services. Emphasis can shift very quickly from value-added network services to being just transport again, due mostly to low storage costs.
The rationale is this: If you can store 500 gigabits worth of information on a home server located by your TV set, why pay a premium to access it from your network provider? I can just download and store the top 200 requested movies and have them immediately available whenever I want, buy them straight from the source (Hollywood) and never need to worry about a movie-on-demand service driven by my broadband connection. This could be totally managed by, say, the Microsoft Windows XP MCE platform, or some other application. The abundance of storage on the local home network will substantially affect a lot of the business cases out there for value-added applications and network services.
Voice over 802.11
Is 802.11 finally the killer app that will make VoIP more widespread? Many believe that within the enterprise it is. Simply put, this technology is the ability to utilise your 802.11 network for voice communications within your campus. 802.11 handsets are beginning to flood the market for enterprises to use in this fashion. Yes, this is interesting news for cost savings and integration without wires, but at first glance might not be overly exciting.
Where it gets exciting is when vendors start to build vertical market specific functionality based on this technology. The implications now become far-reaching. One example of a vendor changing the model for doing business is Vocera, which has launched an 802.11-based hands-free voice communicator for the health care industry. This is an ID badge-sized product that clips to the clothing of hospital workers who often don't have their hands "free" to dial the phone. It's very Star Trek in look and feel and allows for voice-activated dialing within the hospital campus to check records or call for emergency assistance.
Today, these products are focused on workers who would leave their communication device at work, but in the future, networks will allow for connectivity outside of the campus environment through roaming capabilities. We also like the new range of self-discovering 802.11 products that enable you to setup a VoIP presence anywhere by simply powering them on.
This term has caused quite a deal of confusion in the market as vendors in this space attempt to define themselves differently in order to gain market attention. Most simply put, a session controller is the ability to provide routing and control to manage real-time traffic flows between IP networks. These sessions can include voice and/or video traffic. The carriers need to deploy such functionality within their networks to protect their IP traffic security and that of their customers.
When sending this type of traffic between multiple locations, the questions of firewall and NAT traversal often come up. How does an enterprise allow required traffic flows without allowing traffic flows that might compromise network security? Session controllers can help manage this situation. The hardware/software required to enable this functionality can sit in various locations in the network depending upon which vendor selection is chosen. Some have core products, some edge products and some client-based solutions
Current wireless LAN and PAN (personal area networking) technologies like Wi-Fi and Bluetooth face significant throughput, distance and interference issues. For example, even the fastest variants of Wi-Fi (802.11a and g) will typically provide throughputs of about 20 Mbps -- far less than their rated 54 Mbps speeds, and barely enough for a single HDTV signal -- and will often reach only 21 or 24 metres (often not enough to cover an entire home).
UWB (ultrawideband) technologies overcome these shortcomings by using lower-power signals, which operate over an extremely wide frequency range -- as opposed to the higher power but limited spectrum signals used by 802.11 or Bluetooth products. Notionally, the power levels used by UWB will keep it from interfering with other radio systems in the home or office; UWB signals are often low enough in power to get "lost" in the noise for other wireless systems.
Concerns about the true interference issues caused by UWB has kept the FCC from really letting this technology "loose" in the United States, so current UWB systems have a range of only a few yards, but throughput is high enough that UWB could be used to replace cables in computing and home entertainment systems. In fact, UWB throughput is sufficient to replace even high-bandwidth cables, like those connecting HDTVs to high definition set-top boxes. Given the confusion many consumers face when connecting PC systems and home theatres, this application alone could be a huge market.
With full FCC approval and allowed to go to its full extreme, UWB could very well become the wireless infrastructure of choice for all home networks -- voice, video/entertainment or data.
BusinessWeek originally published this article on 14 October 2003.