Enterprise networking is an area that is full of groundbreaking, disruptive innovations. Some of these – such as WiFi and VoIP (voice over IP) have already made a major impact. Here are some others that are starting to shake up the sector – or are expected to do so in the next several years.
The surest thing in networking's future is the eventual upgrade of the standard protocol underlying the Internet. While the Internet Engineering Task Force (IETF) has successfully patched holes in IP version 4 (IPv4) for over 20 years, the explosion of Internet use, particularly in Asia, has raised concern that the 4.3 billion addresses permitted by the 32-bit IPv4 might run out.
Its likely successor, IPv6, has a 128-bit addressing scheme, allowing for billions more addresses. Its packet header has built-in quality-of-service and IPSec encryption, making it more suited to voice, video and mission-critical applications. Auto-configuration and other plug-and-play capabilities simplify deployment of the large numbers of IP-enabled mobile devices which analysts suggest may drive initial adoption.
Japan has mandated support of the new specification by 2005 and China, as part of its $169 million 'Next Generation Internet Initiative', launched the largest IPv6 network across 25 universities in 20 cities at the end of 2004. While the US Department of Defense has committed to IPv6, most Western companies, lacking the addressing imperative, have been reluctant to rip and replace existing networks.
Analysts predict IPv4 will dominate until at least 2009, particularly in corporate intranets, but anyone in the West dealing with Asian companies will have to upgrade their networks to allow the two incompatible protocols to co-exist.
WiMax, based on the 802.16 standard, is an evolution of WiFi, itself a wireless incarnation of the Ethernet wired local area networking standard. With pilot rollouts already underway in New York and Kent, WiMax boasts up to 40Mbps of bandwidth per channel and a range of up to 10km. While this makes it suitable for backbone and last-mile offerings, truly mobile WiMax, offering 15Mbps of capacity within a 3km range, will not be available until late 2007.
Chip giant Intel is one of the technology's chief backers, hoping to repeat its success in building WiFi capabilities into its Centrino laptop processors. Other members of the WiMax Forum include Cisco, Fujitsu, Lucent, Nortel, Siemens, Alcatel, Samsung and Ericsson. But founding member Nokia quietly quit in May 2004, citing the lack of a short-term business case for WiMax. The move did not displease Nokia's biggest customers, mobile operators currently building 3G networks, arguably a rival technology.
WiMax's success is far from guaranteed. Early iterations struggle to work outside line-of-sight of transmitters, requiring costly antennas linked to many internal access points at end-user sites. Contention rates could also be a problem in crowded areas.
Even when established, WiMax's key applications in densely packed cities and rural areas could face competition from long-promised Ultra-wide band (UWB) networks and broadband over power lines, though these too are immature. 3G and 802.22 ('WiFi TV', utilising unused television channels to create regional networks) could either complement or compete with WiMax.
Upgrades to the current 802.11a/g standards for wireless LANs are ratified by the IEEE (Institute of Electrical and Electronics Engineers) with some regularity. The next update is 802.11e, with products shipping in 2005. Among other benefits, this makes WiFi networks more suitable for VoIP. Quality of service improvements allow prioritisation of voice and video transmissions, which have a low latency (meaning messages must be sent and received faster than other data to guarantee quality).
Since early 2004 the IEEE has been working on 802.11n, a more significant development which will offer wider access points and data transfer speeds of 100 Mbps, up to 20 times faster than 802.11b, the first widely accepted WiFi standard. Ratification of 802.11n is expected by the end of 2006.
Mesh networks consist of many interconnected nodes, combined to give decentralised wireless coverage – rather like the way that routers pass on messages over the Internet. Because the network can survive even if an individual node is put out of action, mesh networks are suitable for unstable environments and are cheaper and simpler to manage. Although mobile giant Motorola's recent purchase of start-up Mesh Networks indicates this technology may have a place in future wireless networks, analysts suggest it is still more than five years from maturity.
Even though US networks are still to roll out existing 3G technology, a group of operators and equipment manufacturers (including Vodafone, Siemens and DoCoMo) have committed to developing a 'Super 3G' standard which will enable data transfer speeds 10 times faster than today's 3G networks.
While this is not expected before the end of the decade, it could provide a stop-gap until 4G arrives. Though it will be at least 2012 before products are developed, 4G could allow data to be exchanged between mobile devices around the globe at speeds of 100Mbps and locally at 1Gbps.
In spite of this enduring hotchpotch of wireless technologies, mobile users still need to be able to move seamlessly and transparently between the various networks. Starting off in this direction, a group of 14 companies – including Alcatel, BT, Nokia, O2 and Motorola – published specifications for Unlicensed Mobile Access (UMA), which would allow handover of voice and data calls from 'outdoor' GSM cellular to 'indoor' IP networks, such as WiFi.
Motorola, in partnership with Avaya and Proxim, is soon to release a handset which can move from GSM to WiFi without interrupting a call. Breaking down distinctions in wireless networks is vital to the long-term goal of converging fixed and mobile applications onto one IP-based platform. The 2005 incarnation of this will be BT's 'Project Bluephone', which promises one phone that can be used at home, in the office or at special public hotspots.