Thought Pad

Dr. G. Venkatesh, Chief Technology and Strategy Officer, Head-Product Division, Sasken Communication Technologies Ltd.

What should a lay person understand from the phrase "wireless broadband"? "Broadband" is probably better appreciated thanks to DSL - it means high speed connections enabling richer experience of content on the Internet. Surprisingly, "wireless" is harder to describe - since it represents different uses for different people. In emerging markets with scarce telecom infrastructure, wireless means accelerating telecom service deployment that was hitherto impossible due to the prohibitive cost of laying copper.

In other markets, "wire-less" conjures a world free from all the messy wires we see around our devices. To most, wireless means mobility - it means getting access to voice, data and other services while walking around, commuting by car or bus, zipping along a highway, or on a train; maybe even while on a ship or a plane. To the globe trotters, wireless could mean "roaming" access - get connected to a nearby network, be recognized and provided with services as if you were at home.

The combination "broadband wireless" is even more confusing in its interpretation. For some, it is about getting access to the Internet without having to look around for a Wi-Fi access point. For others, it is about getting DSL -like connectivity even though there is no copper to the home. It could also carry the image of a cellular system on steroids.

It is thus necessary to first demystify broadband wireless. The extent of broadband we can get out of a wireless connection depends on two factors - the amount of spectrum available per user, and the efficiency with which that part of the spectrum is used.

To start with, the amount of usable spectrum is limited, and precious parts of it are wasted - for example, spectrum (around 400 MHz) very suitable for wireless is allocated for terrestrial television, and is unused because most of us get our TV through cable. On the other hand, frequency bands allocated for wireless communication- the 2.4 GHz unlicensed band used in Wi-Fi, and the 2.3 and 2.5 GHz bands allocated for WiMAX - are not very wireless friendly. Nevertheless, it is heartening that more spectrum is starting to be made available by regulators for wireless communication.

Now allocating more of the spectrum for one user means there is less of it left for others. Cellular technology overcomes this by 6 TIMES GLOBAL JOURNAL - 2006 7 THOUGHT PAD Thought pad Dr. G. Venkatesh, Chief Technology and Strategy Officer, Head-Product Division, Sasken Communication Technologies Ltd. reusing the same frequency between users who are sufficiently separated spatially, by locating and associating a user with a cell. In any case, till recently, technology did not exist to handle more spectrum per user - since it was hard to make analog parts work uniformly over a larger frequency band, and since available battery power could not support the higher computational needs of the digital parts. Advances in semiconductor technology have now made both these possible, and we are able to deal easily with 5Mhz or 10MHz per user. One way to get efficiency from spectrum is related to our ability to counter the eccentricities of the wireless channel, which changes with space and time. But this is difficult. A more fruitful direction to get efficiency is through interference management.

Inherent in every wireless system is interference - for example even if two users are allocated different frequencies (FDMA), it is not possible to design their transmitters so that they don't interference with each other. Likewise, two users who are allocated different time slots (TDMA), could interfere when they transmit over different distances. As we saw earlier, cellular systems allow interference in order to reuse frequencies. The brilliance of the CDMA system is in recognizing and embracing interference management - two different users are allowed to interfere by transmitting on the frequency and time slot, but can be distinguished from each other because they use different codes. The design of the cellular system also becomes easier, since neighbouring cells can share a frequency and be distinguished by codes. By using a larger frequency band (5 MHz) for each user, WCDMA carries this idea one step further.

Though we have come a long way with wireless technology, the spectral efficiency of current wireless systems is still quite poor - less than 1 bit/sec per Hz of bandwidth. By using adaptive modulation and better coding techniques, some of the emerging technologies like HSPA and EV-DO improve upon this. But the two most promising technologies that will push the frontier of wireless to the next higher level are OFDM and MIMO. In OFDM, each user is allocated a large frequency band, but this band is subdivided into smaller bands. The system monitors the wireless characteristics of each of these sub bands to push more bits through the favorable bands. More interestingly, OFDM can manage interference dynamically. MIMO looks at the interference issue differently - by using several antennas at the transmitter and receiver, it is possible to send the transmission along the direction of the receiver, thus reducing the interference in other directions. Even better, it is now possible to create several parallel beams through space from one point to another, thus increasing the capacity realized from a single frequency. Using OFDM and MIMO, it is now possible to extract 10 bits/sec or more per Hz.

It is therefore not surprising that nearly everyone is embracing these technologies in their future systems. For example, OFDM is used in one variant of UWB (the technology for broadband BlueTooth), in the long term evolution (LTE) of GSM, in a variant of Wi-Fi, and is the principal differentiator of WiMAX.

Given that wireless broadband is now feasible; will it become commercially available on a wide scale soon especially in markets like India? Sadly, the answer is no. There are two reasons - one is cost related and the other is business model related.

The cost of a network depends directly on the number of cell towers required to get the required coverage and capacity. On this parameter, all the broadband wireless technologies (HSPA, EV-DO or WiMAX) score very poorly. In spite of all the hype surrounding broadband wireless, it is not possible yet to create a bandwidth in Mbps at distances over 1 Km unless a higher power transmitter is used, or it is restricted for fixed wireless in which the terminals are equipped with external outdoor antennas. The former is probably not a good solution - higher power transmitters means more interference not less, and could lead to health related issues. On the other hand, external antennas increases the terminal cost substantially - only rich rural households or enterprises will be able to afford it. Without range, broadband wireless will end up being deployed only in dense urban locations. Even here, the presence of heavy concrete structures in buildings (in India) may lead to disappointing bandwidths indoors. Thus, increasing the range of broadband wireless systems remains the single most challenging technological issue that will determine its success especially in emerging markets like India.

From the business model context, we note that operators are struggling to find new sources of revenues. Voice revenues are declining. Data has not taken off. Will wireless broadband be the angel which will save the operators? There is a danger lurking here - with broadband wireless, it is possible to push voice over packets (VoIP) over the broadband link, which means that the flat rate business model for broadband will cannibalize the lucrative usage based fees for voice. Operators could thus have the genuine fear that this angel in disguise could transform into a monster that will devour their existing business model. Consequently, they might prefer selective rather than wide scale deployment.

So then what is the way forward? There is yet another interesting alternative motivated by the following question - what kind of services will get deployed once broadband wireless is available? Video seems to be a definite bet. A lot of video content is consumed in countries like India. But a much better way to deliver video content is mobile TV, which combines the best of broadcasting and cellular technologies. In fact, mobile TV uses the efficient 400MHz spectrum for broadcasting video. But who will drive mobile TV? Will it be the cellular operators or the broadcasters, or both? Can they learn to co-operate?

In summary, technological advancements will make broadband wireless a reality in a few years. But unless wireless broadband can be delivered over much higher distances, we are not likely to see large scale commercial deployment of wireless broadband. Surprisingly, this could actually suit the operators, who may be more comfortable with only selective deployments of wireless broadband. Meanwhile we could well see the most interesting broadband content (video) delivered through an alternate broadcast channel through mobile TV. Maybe this will kick start the search towards genuinely new content and applications that will then drive the deployment of broadband wireless.