Yesterday,
DS2, a leading provider of chipsets for powerline networking devices, hosted a demonstration of its next-generation equipment for the press and financial analysts. The hardware was impressive, handling data at sustained rates of well over 400Mbps while maintaining backwards compatibility with earlier generations of the company's hardware. But, while powerline networking is a promising solution for a number of problems, there are some substantial challenges to its adoption by the broader market. Built for speed
DS2, based in Valencia, designs the processors and creates integrated chipsets that enable OEMs to produce networking solutions; they don't actually sell end-user hardware themselves. The previous generation DS2 chipset, introduced four years ago, enabled communications at 200Mbps over standard power lines. That generation has been widely deployed by European telcos, and displaces earlier equipment in consumer hardware.
The new generation doubles the bandwidth to 400Mbps a second. Even after control, error correction, and quality of service information are subtracted, the end users will have access to well over 200Mbps of usable bandwidth. That's double 100-BaseT Ethernet and better than 802.11n can do under most conditions. The demonstration hardware came in a carrying case that was roughly the size of an unabridged dictionary. It was shown handling data transfers at 230 Mbits/s between two laptops, happily sending its signal through surge protectors and power strips.
DS2's test equipmentHopefully, production units will be more compact than DS2's test equipment
DS2 has ensured that its hardware is backwards compatible with earlier products, and showed the new device talking to current consumer wallplug adapters at about 100Mbps. In contrast to wireless solutions, where the presence of earlier hardware pulls the entire network down to its level, each device on a powerline network negotiates its connections independently. Thus, a 400Mbps device can communicate at full speed with whatever generation hardware is present.
The presentation emphasized other advantages over wireless. 802.11 hardware constantly has to readjust its signal and contend with environments (such as New York City, where the demo took place) with many other wireless signals. As a result, data transfer using 802.11n was erratic; in contrast, the powerline transfer shot up to full speed and stayed there. DS2 also implied that differences in the underlying technology meant that powerline data transmission had fewer limitations for future bandwidth increases—they have a lead over wireless, and expected to keep it.
DS2's hardware will also come with some niceties on the software side. 256-bit AES encryption will be handled on the fly, and won't limit transmission speeds in any way. Pairing devices for encryption is vaguely similar to the process of pairing a phone with a Bluetooth headset. The chipset will also host a lightweight, POSIX-compliant operating system that will allow remote management; they demonstrated it hosting a web server that displayed performance data. Although the hardware is primarily being sold through its use of powerlines for data transmission, it will also work over coaxial cables and twisted-pair wiring, adding to the flexibility of its deployment.
Who needs the speed?
It's pretty easy to view 400Mbps as overkill, but the company's CEO, Jorge Blasco, did his best to argue otherwise. He pointed out that, four years ago when the 200Mbps hardware was introduced, everyone viewed that as excessive. Due to market realities, however, the 200Mbps equipment has only recently reached the point where it represented the majority of the powerline hardware sold. They want the hardware on the market as soon as possible, so that it's common and cheap by the time that bandwidth needs catch up. Besides, Blasco said, "speed is never high enough if you can deliver it at the right cost."
Planning for the adoption curve: NPD Group data shows that it takes some time for new equipment to displace earlier generations.
DS2 also has some pretty clear ideas of what will drive future bandwidth needs: delivery of HD content and IPTV. Chano Gomez, one of the company's VPs, expects HD movie downloads "very soon," and highlighted IPTV self-install kits offered by BT that include powerline hardware. According to DS2, the new hardware will allow IPTV providers to send 10 HD channels down the pipes simultaneously.
Gomez also suggested that 400Mbps hardware would be even more critical once content makes its way inside the home, as home networks are going to need roughly three times the bandwidth of their connection to the outside world. That's because of centralized storage devices, such as HD-DVRs and NAS boxes, that are going to need to be able to send HD streams to several locations in the house while simultaneously supporting computer backups and internet access. Suddenly, 400Mbps doesn't seem like overkill.
Remaining barriers
So, why aren't we all using powerline network equipment? In Europe, there are nearly 400 million users, but they are very unevenly distributed: over half are customers of Telefonica in DS2's backyard, Spain. In the US, regulations can vary from state to state; in some cases, power companies are prohibited from using their wiring for anything other than the delivery of power. On the national level, the FCC already considers powerline broadband to be just another information service, so the states appear to be the big hold up.
For the home user, cost of equipment has been a major factor. 802.11 hardware is good enough for most common uses, and its volume of sales and bundling as standard hardware have caused prices of even the most current equipment to drop rapidly. A set of four powerline adapters, in contrast, can run well over $100. That situation is unlikely to change until bandwidth needs outpace 802.11's capacity. Performance has been disappointing as well. Some of the recent HomePlug AV powerline networking equipment we've tested at the Orbiting HQ has failed to impress, with speeds lagging 802.11g, let alone 802.11n.
Worse still for both consumers and ISPs are compatibility problems among the equipment from the three main providers—the DS2 execs were happy to point out that some of their competitors make equipment that's not even compatible with their own earlier generations, much less anyone else's. DS2 is trying to work with the IEEE to define a powerline standard, but the proposal currently under consideration would allow different "flavors" of the standard that remain incompatible. As things stand now, however, buying equipment from different companies can be a crapshoot.
Posts
Powerline Broadband Hits 400Mbps
ReplyDeleteSubmitted by David A. Utter on Thu, 11/15/2007 - 11:06
The FCC isn't standing in the way of the local utility company rolling out broadband service over powerlines; state governments and the lobbying by competitors have obstructed this.
Ham radio operators have opposed broadband over powerlines (BPL) for a few years, as they cite interference from BPL with their hobby. Compared to the deep-pocketed cable and telecom industries, the hams are a small ripple in a multi-billion dollar pond.
Rather than celebrating the advances made in BPL, most recently demonstrated by the firm DS2, those who could benefit from the service will have to watch from the sidelines as the impact of lobbying against state governments has frozen deployment of the service.
An Ars Technica story on DS2's latest demonstration of its BPL chipset that BPL can hit speeds of 400Mbps. Once other factors have been applied, their chipset can support hardware delivering 200Mbps, double that of 100-BaseT.
This useful commodity could be delivering true broadband service not only to underserved rural markets, but urban areas where rewiring a skyscraper for fiber would be a monumental expense. Big buildings and bucolic farmhouses have something in common: existing electrical wiring.
Since this would be a municipal operation, prices would be far less for service than existing competitors who won't close the last mile to customers to provide service in the first place. That has the telecom industry lobbying against municipal BPL in several states, successfully in many cases.
It's not just the telecoms who have reason to fear BPL finally fulfilling years of promise. Cable television can be delivered to a peering point and pushed over BPL; it's just ones and zeroes like other data, but again, at a potentially much lower price than the local cable monopoly charges.
Interference with FCC-approved ham radios has been the major point of contention. The well-intentioned delays caused by the hams have allowed cable and telcos time to head off promising BPL initiatives long before they can be considered.
People shouldn't put up with broadband service that underperforms alternatives, especially when considering just how much they pay for it. BPL needs a solid standard, which the IEEE has started to explore; that standard needs to consider how to protect ham radio frequencies.
Compared to the political problems and the money the entrenched interests can throw at politicians, we hope such a standard emerges quickly.