indepth
WIRELESS

As technology becomes increasingly driven by the demands of convergence, wireless is at the heart of a genuinely symbiotic relationship.

An ever-increasing array of multimedia-capable devices, driven by breakthroughs in carrier technologies, hardware and software are dovetailing to bring everything in the wireless world together. To say that the medium is the message has never been more appropriate. With the social media phenomenon rapidly morphing into Enterprise 2.0 for the business world, consumer and business markets look set to bleed into each other to get the most out of what’s on offer. This in itself brings huge pressure to bear on wireless providers as they grapple with the ever-present quality of service (QoS) challenge.

 

CONNECTIVITY, PURE AND SIMPLE

The continued growth of mobile social networking and location-based applications/services is adding to a business mix where users seek increased collaboration functionality, jockeying for bandwidth with millions of users simply seeking to socialise on a global scale. Significant growth in mobile Internet services is proving something of a double-edged sword for wireless providers as they strategise to avoid becoming mere ‘dumb pipes’ carrying the kinds of services that make massive profits for other enterprises. Providers are increasingly seeking to offer services of their own – all while trying to maintain the speeds necessary to make these services useful at all.

  It’s something of a chicken-and-egg situation; trying to beat the service developers at their own game while avoiding rolling content and products out for their own sake is a fine line to walk. As Fierce Wireless editor Sue Marek told delegates at an IFA press conference in April 2009, developing a product in the hope that end-users can be “educated” into finding it useful is not a good idea: “That’s called product failure.” Marek’s assertion that “whoever owns the connection will own the relationship with the customer” illustrates the fine balance that those operating in the wireless market will have to tread in the wake of device proliferation. As the Yankee Group points out, new entrants, all-you-can-eat service plans and operator bundles are changing the business of access. With cash-strapped consumers cherry-picking across mobile, wired and wireless networks to create their own version of a mashup, the future is set to head in the connection of pure-and-simple connectivity, regardless of carrier technology – turning ubiquitous connectivity into sustainable profit is the next challenge.

 

Blurring the communications lines

Wireless and cellular networks have, between them, the potential to offer pervasive connectivity at 50Mbps, converging to blur previously easy-to-see lines. When talk of WiMAX or Long Term Evolution (LTE) is as much about your new smartphone as it is about more traditional data transfer, you know we’ve come a long way down the convergence road. Dr Sanjoy Paul from the Convergence Technology Lab at Infosys told delegates to the MIT EmTech India conference in 2009 that Internet users will likely reach two billion by 2012, and “50% of those are going to come from wireless mediums. The total world population by 2013 is going to be 6.8 billion and around 80% of these are going to be mobile users. So demand for bandwidth is going to increase manifold.”

Hardly surprising, then, that spectrum – and the allocation of it – has come to the fore as an issue in pretty much every market around the world. And rather than become embroiled in a discussion of whether WiMAX, 4G or LTE is at the core of the wireless future, big hitters such as Intel and Motorola are hitching their wagons to a variety of carrier technologies in order to ensure that their developments will ride any wave. With Cloud Computing creating a lot of buzz through its promise of taking tools and applications off the desktop and into a hosted “cloud” from where they can be accessed on demand, it’s worth remembering that none of this – or many of the other technologies we now take for granted – would be realistic if our wireless networks weren’t up to the task. In a wireless world, the network is the computer.

 

STANDARD MIX

Wireless data communications can be broadly segmented according to network type, among them wireless personal area networks (WPAN), wireless local area networks (WLAN) and wireless metropolitan area networks (WMAN). Each segment typically supports a variety of technologies to connect users.

The Institute of Electrical and Electronic Engineers (IEEE) establishes the standards for each offering, providing a centralised development forum to ensure interoperability. Apart from dealing with potential problems such as interference, standards are also crucial to achieving security levels equal to that of a closed, firewalled network. Among the wireless network specifications are the following:

802.15: Short and personal With a range of only a few metres, this standard governs technologies such as Bluetooth, personal area networks/wireless devices in close proximity to each other such as PDAs, cellphones, pagers and other consumer electronics. In terms of where the future might be going, this standard has a task group working on Body Area Technologies (BAN) with a goal of low-power, low-frequency short-range wireless.

802.11: Wi-Fi’s evolutionary alphabet Thanks to a lengthy (and, indeed, ongoing) evolutionary process, 802.11 has chomped its way through a fair amount of the alphabet. Early versions were 802.11a (about half the speed of Ethernet) and 802.11b (equal to standard Ethernet) but we’ve moved on through 802.11g (offering around 54Mbps using the longer-range 2.4GHz frequency to arrive at the latest iteration: 802.11n. Publication of the standard is expected in November 2009, but many businesses have already begun switching following the Wi-Fi Alliance’s certification of products that conform to the 2007 draft on the standard. 802.11n builds on bandwidth support through the use of MIMO (multiple-in, multiple-out), which allows it to use multiple signals and antennas – ultimately, it should readily-support rates in excess of 100Mbps.

Signal intensity, meanwhile, sees n offering superior range to its previous Wi-Fi inceptions. All equipment will be backwards-compatible with 802.11g. On the down-side, there are concerns that multiple signal capacity may interfere with related 802.11b/g networks. 802.11s covers mesh networking (more of which later). 802.11u is underway but not expected to be made a formal standard until March 2010, when it will address the issue of improved interoperability between networks, offering improved-quality handover and roaming. The main application for 802.11r – which accommodates hand-offs aboard moving vehicles – is likely to be VoIP capable of working with or without cellular networks.

802.16: Broadcast connections, broadband wireless These typically involve network connections from single base stations to many devices over a large area – wireless metropolitan area networks. To most people, however, 802.16 means WiMAX (more of which later) which, although far from ubiquitous, is nonetheless following a fairly clear growth path. In its mobile form, WiMAX acquires the letter e – 802.16e, which is widely viewed as closely comparable to the developing “LTE” mobile wireless standard. For this reason, many organisations working towards WiMAX are using the 802.16e standard for both fixed and mobile roll-outs. Fixed WiMAX is known as 802.16d.

The standard for 802.16m, – is currently at the draft stage, with an IEEE working group reporting regularly on progress. The aim is to support air-transfer rates of 1Gbps for fixed and 100Mbps for mobile. As with 802.11g, MIMO will play a central role. 802.16m isn’t part of the WiMAX standard but the IEEE has suggested that, not only will it be compatible with 4G networks; it will also play nicely with WiMAX.

 

LOCATION, LOCATION, LOCATION

Wireless networks are location-based. To get optimal results, it’s necessary to identify each connected device, determine its location and then track of it moves from one wireless access point (WAP – confusingly the same acronym as wireless access protocol, which it is not) to another. This must be achieved without creating unrealistic burdens on network traffic or sacrificing key requirements such as security, user authentication and encryption.

Wireless networks are prone to unpredictable peak loads. For example, the drop in service quality experienced when every wireless user in one building accesses the same lone WAP during a meeting. Unlike wired connections, where the real bottlenecks lie in server capacity, wireless networks often find their point of pain at the WAP. Consequently, each WAP should have mechanisms in place for apportioning user connections, based on factors such as usage levels or proximity.

Location also influences security. With no physical perimeter, unauthorised users outside a premises can, theoretically, join the network. If security is inadequate, IP addresses and network SSIDs can be harvested, allowing crackers to set up their own WAP, forcing legitimate users to unknowingly connect to it, exploiting the reality that wireless devices automatically look for the nearest or most available (in terms of traffic) access point. These “evil twin” set ups can allow unauthorised users to access the same network services as authorised users, with predictable consequences. Network admins are only too aware that wireless networks can be seen as “dirty”, and so block off access to the rest of the network. At both hardware and software level, features are included that support network access authentication, usually at layer 2 – the link layer – of the network.

 

BLUETOOTH TO GROW TEETH?

By far the best-known wireless personal area network (WPAN) technology, Bluetooth allows users to connect different devices and exchange data between them. As the business and leisure worlds dovetail towards a mobility-driven landscape, the ability to connect and synchronise devices becomes increasingly important. With the technology ramping up to meet changing usage patterns, Bluetooth is edging away from a largely cable-replacing technology, allowing groups of users to share data even when there is no LAN or Internet infrastructure available.

Despite massive popularity with end users, many commentators felt that Bluetooth was losing relevance in a 3/4G, Wi-Fi-enabled world. The arrival of Intel’s Wi-Fi PAN, which allows Centrino-based devices to connect to peripherals (including the iPhone) looked set to steal much of Bluetooth’s thunder – until the latter launched version 3.0+HS, which is capable of using the Wi-Fi 802.11 standard to transfer larger files. The H” stands for ‘high speed’ – and the Bluetooth Special Interest Group (SIG) is claming transfer rates of 22-26Mbps; products are expected to be widely available towards the end of 2009. The key advantage is that devices using ad-hoc connections can transfer data at high speeds without needing to access a Wi-Fi network or have a USB cable.

Bluetooth 3.0 also claims enhanced power consumption features through better battery life – which can also reduce the chances of being “bumped” off a connection if the device is moved to a pocket or bag. According to Gartner research, Bluetooth is a “short-term” technology “for businesses and IT organisations to watch during the next three years,” citing 3.0’s capacity to “deliver major improvements in power consumption and features with device shipments starting in 2010. Low-energy (LE) Bluetooth will also enable new sensor/ peripheral devices and business opportunities.”

Fans of Bluetooth point to lower power consumption and production costs compared to Wi-Fi while critics have long-since pointed to what they claim are over-inflated speed reports and security issues – so-called “man in the middle” attacks in which users inadvertently pair with the wrong device. Improved security via features such as multiple-digit pass keys and encryption has eased the pressure on Bluetooth in this department, however.

 

 

NEXT-GENERATION WI-FI

Wi-Fi (Wireless Fidelity) is the term used to refer to any product or service operating under the 802.11 standard with networks carried in the unlicensed 2.4 and 5GHz bands. With throughput of 11Mbps (802.11b), 54Mbps (802.11a/g) or 100Mbps (802.11n), Wi-Fi-enabled devices can send and receive data from any location equipped with access points that transmit a signal to devices within a range of around 100 metres. Despite growing up on the scruffier fringes of the wireless world, where it shares unregulated spectrum with a motley crew of devices, high adoption rates and a reduction in the cost of enabled laptops have ensured Wi-Fi’s success as a technology, with some of the biggest names in IT throwing their weight fully behind it – most notably Intel.

 

 

WHITHER WIMAX?
  WiMAX has enjoyed only a brief period of buzz in between its status as Next Big Thing and its new position as a technology whose future is hotly disputed. Despite enjoying significant success in Asia, observers further afield are giving more than a few column inches to pondering WiMAX’s relevance in the face of emerging 4G/LTE. WiMAX research organisation Maravedis has regularly stated that it did not “expect WiMAX to become a 3G killer in the near future. This remains true in light of the recent technical and commercial wins by LTE.” Be that as it may, Maravedis reports that more than 1.2 million 802.16e-2005 (802.16e is the unofficial standard for mobile WiMAX) compliant devices and chipsets supporting mobility were shipped in 2008; in the fixed WiMAX device category (802.16d) reporting a relatively-healthy 880,000 shipped.

 

An African future?

Having taken rather a long time to begin rolling out in South Africa, WiMAX now looks like it’s sitting on a tipping point – its status as a last-mile delivery solution in a country with infrastructure challenges indicates that it certainly has a future in Africa. Whether that future involves a fulfilment of all that WiMAX promises remains to be seen, however – delays in awarding licenses, disputes regarding spectrum allocation and relatively-limited roll-outs have been something of a damp squib to date.

Last year’s bone of contention – the requirement that those bidding for spectrum be 51% black-owned companies – has been diluted significantly by the recent reduction in that requirement to 30%, allowing smaller operators to get their slice of the WiMAX action. Nonetheless, it’s fair to say that South Africa lags behind the rest of Africa when it comes to roll-outs. According to the World Wide Worx report “WiMAX in SA 2008:

Year Zero”, the only companies capable of appreciating the true impact of the technology are those that have already deployed it. The hurdle in the way of such full appreciation is, according to World Wide Worx, the lack of sufficiently serious competition to drive roll-out which, according to MD Arthur Goldstuck, has kept expectations artificially low, feeding perceptions that the technology is not capable of delivering. The report found that most companies are using Telkom’s scaled-down WiMAX offering (available only in areas where ADSL fixed-line services are scant) – and therefore only achieving ADSL-type speeds, which at a current maximum of around 4Mbps is far below WiMAX’s theoretical capabilities of 70+Mbps.

 

Supercharging the last mile

Because 802.16 operates over a high frequency, speed of 70-100Mbps over a 122km radius are – theoretically, at least – possible. Even the more realistic speeds of 4Mpbs or, wonder of wonders in SA, 10Mbps are well ahead of what average users locally are experiencing. As indicated above, local perceptions of WiMAX revolve largely around its role as a last-mile delivery solution complementing Wi-Fi and 3G/HSPA solutions, getting into areas with little physical infrastructure, be it copper or fibre optic. On this point, at least, all the local players are singing from the same hymn sheet, particularly Telkom and Neotel. Almost all players in the SA market are offering fixed-line services over 802.16e networks. Telkom was the first to market back in June 2007 but by mid-2009 reported that it was undertaking a review of its options. Towards the end of August 2009, the organisation announced that, contrary to rumours that it was abandoning WiMAX to focus on W-CDMA, it would continue with WiMAX provision in areas where it makes sense to do so – i.e. those with little or no copper infrastructure. Neotel, meanwhile, has announced WiMAX services aimed squarely at the SME sector, where it will compete with the likes of iBurst/Vodacom for a share of the small business/high-end consumer market.

WiMAX works in much the same way as cellular networks, using transmission towers to pick up and relay signals across a wide area. At the end of the chain, receivers in the form of hubs, PCMCIA cards or processors pick up the signal, bringing wireless data transmission to the end user. With the necessaries in place, it’s viewed as a genuinely low-cost option; price paid for Hz for WiMAX spectrum is as much as 1000 times lower than that paid for 3G, according to Maravedis research.

When it comes to convergence, WiMAX certainly looks the part – potentially massive range, speeds that make triple/quadruple play a reality and more efficient use of spectrum. Unlike Wi-Fi, which operates using fixed spectrum width, which limits it, WiMAX facilitates the extremely close, concurrent running of signals along wireless channels, giving users more bang for their bandwidth by carrying more data through even narrower pipes.

WiMAX can also take the Wi-Fi channel and sub-divide it, allowing for many more users at a time. Because 802.16 networks use the same logical link control (LLC) layer as other LANs and WANs, it can be bridged and routed to them. It also supports multiple physical layer specifications, making for a standard that readily adapts and evolves to a variety of different requirements, allowing equipment manufacturers

 

 

 

 

WiMAX in the long term

According to Telkom’s Gerrie Opperman, “convergence will happen at 4G, which is in WiMAX’s evolutionary path.” International trends indicate that WiMAX is certainly playing a significant role in the mix, but many commentators are nonetheless convinced that the imminent arrival Long Term Evolution (LTE)/4G technology will somehow sidestep it, leaving it behind. Maravedis suggests that, to date, many WiMAX providers have not implemented LTE because the technology is still to far down the road to present a realistic option and many prefer to make the most out of the spectrum they currently have. Add to this a financial crisis that caused most organisations to re-think their investment strategies and the fact that bighitters such as Nortel, Nokia and Motorola have either “re-adjusted” or otherwise abandoned their WiMAX strategies in favour of LTE R&D and you can see why some believe that WiMAX is on a downward curve. Others are quick to point out that WiMAX’s failure to gain meaningful traction in Europe thanks to spectrum disputes doesn’t necessarily translate into international failure – the technology continues to gain ground in Asia and Africa, for example.

Both WiMAX and LTE are 4G networks driven by an emphasis on shifting data packets over IP. LTE has its roots firmly in the cellular department, with WiMAX starting out at the opposite end of the scale. This latter point probably accounts for much of the negative perceptions of WiMAX – in a world increasingly dominated by mobile connectivity, some just cannot see the technology being able to up its game to carve a niche for itself. LTE is promising delivery of downlink speeds of 3-10Mbps with uplinks at 1.5-3Mbps, depending on how much bandwidth the carrier can offer. Many commentators note that LTE presents a gilt-edged opportunity to settle on a single standard, moving away from 3G’s CDMA-GSM spats and the dual-mode devices that must follow in their wake for those seeking to roam.

WiMAX, meanwhile, offers 10-15Mbps that carriers can slice up/down any way they choose. And while LTE is not expected to hit the kind of critical mass required for market domination before 2012, WiMAX is already here in both fixed and mobile formats, which should see it retain some market share, not least in developing markets (back, again, to the seemingly-inevitable association of the technology with plugging infrastructure gaps). Perhaps the most useful way of looking at the scenario is from the point of view of dovetailing technologies. Many WiMAX technologies and settings have found their way into LTE offerings, with some commentators going as far to suggest that the two standards be merged. With near-identical offerings, it seems inevitable that the two technologies will square up for another iteration of the VHS-Betamax/BLU-Ray/HD DVD smackdown.

 

MONSTER MESH: 802.11S

Sometimes called ad-hoc networking, mesh networks are self-organising, capable of assessing what’s going on with other nodes in their vicinity in order to work out the best route for the data they carry. A popular analogy used to explain how mesh works is the cocktail party: you arrive and, not seeing anyone you know, introduce yourself to a few people, working your way around the room until you finally see the person you’d arranged to meet and negotiate your way through the crowd to make contact with them. If your friend hadn’t showed up, you could still talk to others and make contact with people throughout the room. And if a lot of people show up, you can extend the party out into the next room or the garden – if you’re still looking for your friend, you can ask other people if they’ve seen her, and on you go. That’s mesh: self-organising, scalable, self-healing.

 

What’s so good about it?

Mesh networks allow users to build particularly strong networks thanks to the availability of multiple routes. Even if several nodes lose power or malfunction, there are still options available before the network fails completely. Each node acts like a router and each one is capable of “talking” to another, eliminating the need for everything to be routed back through a base station – making scalability a genuine possibility as you don’t need to install new base stations to extend the network. With each node acting as a router for others, data is moved around between them in a process known as “hopping” – a key differentiator between mesh and a traditional hub-and-spoke network.

Wireless mesh is well suited to extending bandwidth reach at a relatively low cost, offering scalability and flexibility to boot. Particularly well suited to metropolitan networks, it’s no surprise that they’ve proved popular in cities such as Panama City, Brooklyn and Mumbai. For business, wireless mesh can facilitate a shift from capex and into the more palatable realm of operational expenses on the network-expansion front. Thanks to the flexibility, businesses can pay as they grow, as scaling is relatively painless and fast. It’s a lot less expensive to run power cables to nodes that it is to roll-out data cabling; the IT department may only need to connect, say, every 10th node to the physical network, allowing the rest to travel through the air.

 

For every silver lining there’s a cloud

On the downside, it’s been argued that mesh technology’s origins on the battlefield (it was first developed by the American military) lie at the heart of its negative points. Yes, they’re relatively easy to set up but, as with many technologies designed with military applications in mind, their throughput and range are limited. A key complaint is that once traffic is back-hauled over more than four nodes, throughput declines to a lowly 7Mbps, raising questions over its capacity for wide-scale metro networks or handling voice and data-heavy networks.

On a positive note, nodes capable of easily-handling a throughput of 35Mbps over multiple hops are now available. A further possible speed bump comes in the form of the 802.11 radios mesh uses for backhaul; where several radios operate in the same channel, competition for bandwidth will ensue. This effect is amplified as the number of hops increases. Furthermore, the fact that mesh networks aren’t “pure” 802.11 can cause headaches on the WLAN integration front. Standards and algorithms exist to work around these issues, however. Perhaps the biggest opportunity for mesh likes in its potential application in the field of pervasive computing, which may one day become its killer app. Small sensors can be embedded in anything from sunglasses to electronic devices and clothing, which could be connected to smart networks over mesh. Of course, the possibility of quality metropolitan networks is killer app enough for many...

 

AN UNSTRUNG FUTURE

As more and more carriers scramble for customers, issues around licensing and spectrum look set to occupy the minds of those aiming for pole position in the wireless space. The key challenge – as with mobile – is capacity to carry data-rich applications to users with low tolerance for poor quality. It’s no longer enough to sing the praises of mobility, previously the key selling point for many wireless technologies. That particular novelty has worn off and is more or less expected as standard these days – users now want increased speeds, better quality, more services and lower costs.

As seemingly everything moves to wires-free, the lines between traditional service providers and those in the cellular world have blurred beyond recognition, forcing well-established organisations to re-think their strategies if they are to survive in a new environment in which users are less likely to worry about the technology that’s carrying their data as they are about seamlessly accessing and sharing that data. As the number of devices from smartphones to Netbooks and beyond continues to proliferate, those organisations that can keep pace will be the ones capable of looking at a platform agnostic future in which their role as provider has had a serious re-think.