Communication big bang

Why is there a trend in going for tetherless/wireless access technologies? The obvious reason is that these technologies can do away with installation and maintenance cost of wireline systems and facilitate different levels of mobility. However, unlike wireline, the radio medium is a shared entity between all users within the coverage range of the central hub (cell site) with which the customer premises units communicate.

This adds flexibility. But, there is also need for regulatory measures and sophisticated techniques to control mutual interference and overcome radio fading.

In a simple point-to-point mode, multiple users can be multiplexed, using different frequencies (FDM), different time slots in digitized timeframe or different spreading codes (CDM) in the same frequency bandwidth and timeframe. Variation of the same techniques is also used for multiple accesses, so long as the users are within the coverage range of a hub (cell site) for traditional use and each other in the case of ad hoc operation.

Thus, we have Frequency Division Multiple Access (FDMA) with a bank of frequency carriers, Time Division Multiple Access (TDMA) with a set of time slots, and Code Division Multiple Access (CDMA) with different spreading codes. Each of these technologies also makes use of sectorial antennae to reduce interference between adjoining sectors and improve capacity

Based on complexity, all the first generation (G1) systems in 1970s were analog, and used FDMA mode of operation which yields poor spectrum efficiency due to requirement for large signal-to-noise (S/N) values. This was improved in the next generation (G2) mobile systems in 1980/1990s, which used digital techniques to bring down S/N or Eb/N0 (Bit Energy/Noise Density) requirement. A popular example of TDMA system is the GSM as used in many parts of the world.

Late 1990s also saw the emergence of CDMA- based G2.5 systems, which further improved system capacity, not only due to lower value of Eb/N0, resulting from signal spreading and collapsing, but also as a result of common frequency reuse among various sectors and cells. Among other benefits, the technology facilitates better utilization of communication resources with multi-rate coding, and is the basis of IS-95/1xCDMA standards.

Due to its various advantages, the CDMA technology has also been adopted for the three major 3rd generation (3G) standards viz. 3G Partnership Project (3GPP) based UMTS/W-CDMA, 3G Partnership Project 2 (3GPP 2) based CDMA-2000 and China led TD-SCDMA. The CDMA-2000 standard makes use of 2G (1xCDMA) technology for voice, but for broadband data communication, new channel architecture has been evolved leading to Evolutionary Data Optimized (1xEVDO / 3xEVDO) system. T

he same is the methodology for UMTS/WCDMA, which uses wideband but conventional CDMA technology for voice. But, high-speed data is facilitated using High Speed Packet Access (HSPA), a combination of High Speed Downlink Packet Access (HSDPA) and High Speed Uplink Packet Access (HSUPA). The new 3G standards also allow Adaptable Modulation Coding (AMC) and Hybrid. Repeat Request (HARQ) techniques for better system throughput in addition to IP Multimedia Sub-system (IMS) for session-based channel utilization and control. Many of these techniques are also exploited in WiMax and WiBro standards.

The WiMax and WiBro standards are both based on a specialized variation of FDM/FDMA technologies in which sub-carriers are orthogonal leading to Orthogonal Frequency Division Multiplexing (OFDM)/Orthogonal Frequency Division Multiple Access (OFDMA). Though the technology finds its reference in early literature, the implementation had to wait until powerful microprocessors became available. The technology also facilitates better control on interference allowing use of advanced modulation schemes (say 64QAM) for improved spectral efficiency.

Standardization and Certification

Somebody has compared telecommunication service providers as a pack of animals that always move in herds. There are various reasons for it, least of which is user/operator confidence. A well recognized standard also leads to cost reduction by volume of production and competition in addition to some measure of customer protection against bad suppliers.

The first question with any standard is on maturity of technology. The reliability of OFDM technology can be judged from large-scale operation of a number of OFDM-based standards. This includes Wireless Local Area Network (IEEE 802.11a, 802.11g), Wireless Personal Area Network (IEEE 802.15.3a, also known as Blue Tooth) and Broadband Wireless Access (IEEE 802.16d/802.16e, also known as WiMax and IEEE 802.20 also known as WiBro). The OFDM technology is also used for cable-based ADSL/VDSL (ITU G.992.1), Fiber based radio over fiber, digital audio broadcasting, digital video broadcasting, and other home networking applications.

It is also important to note that 3GPP has sought the adoption of WiMax in its long-term evolution plan. On the other hand the WiBro evolution includes 1x Rev–C/MediaFLO leading to ultra wideband.

The WiMax stake holders have also constituted WiMax forum which has detailed its roadmap for development of accompanying standards including network/service management, operation in license free band, and ad hoc networking. The latest among these is IEEE 802.16-m study group for study of IMT-advanced (4G) requirements, which are likely to be based on OFDM.

What's so Special about WiMax?

While all these systems for broadband data transmission support hybrid automatic request, scheduling based on mix of fairness/throughput, virtual soft hand to avoid loss of system capacity on multiple channels, and power management to reduce interference and maximize throughput, the WiMax standard supports following additional advantages: tolerance to multi-path and self-interference; scalable channel bandwidth (Rel-e) at session level using MAC layer; orthogonal uplink multiple accesses by allocation of different groups of sub-carriers to different uplink users; support for spectrally-efficient TDD to take advantage of greater downlink to uplink capacity requirement; frequency-selective scheduling (Rel-e) by selection of sub-carriers based on instant channel response; fractional frequency reuse (Rel-e) by use of common frequency at center and fractional frequency at edges to maximize system throughput; improved AMC and error correction techniques by better resource allocation in time (slot) and frequency (tone) domain; fine quality of service (QoS) by resourceallocation at session level; and advanced antenna technology (optional for Rel-d, mandatory for Rel-e).

It is easy to notice that WiMax is cost effective for broadband data transmission as illustrated in following comparison with SIMO/MIMO and 1:1 or 3:1 ratio of downlink to uplink throughputs

The relevance

Today, India has emerged as the fastest growing telecommunication market with lowest Revenue Per Unit (RPU). It is also poised for an explosive growth in Internet connectivity with fair share of broadband wireless access technologies to meet Capex and Opex requirements commensurate with RPU. Obviously, the issue is closely tied up with efficient use of spectrum, which is limited.

The discussion in previous sections has clearly established the superiority of WiMax over EVDO and HSPA based 3G technologies for data application. In future, this may also be considered for voice centric traffic based on VoIP technology, as available with push-to-talk in 2G systems, but it would be unwise to ignore advantages of current technologies (2G) for voice centric traffic. The only exception may perhaps be in rural areas where the demand is likely to be data centric based on tele-education, tele-information and tele-entertainment needs.

The same concept is used in development of EVDO (in CDMA2000) and HSPA (in UMTS/WCDMA) channels for current 3G products. The question is whether we can leap frog this development and move straight to broadband wireless access technologies as currently being discussed by 3GPP and 3GPP2 associations.

As quoted by an analyst firm, Maravedis, in August 2007, a similar initiative has been taken in Korea by KT Freetel with more than 20,000 WiBro customers added in the last three months out of planned subscriber base of 200,000 by the end of year 2007.

The same appears to be the philosophy of Sprint–Nextel in USA, which has announced an investment of $ 3 bn for a WiMax-CDMA dual network. This is expected to become operational by the end of 2007 with participation of major electronics vendors like Motorola, Nokia, and LG Electronics-using chips from Intel, Fujitsu, and Texas Instruments.

Can this Happen in India?

Yes, based on Maravedis report of June 2006 indicating WiMax subscriber base of 13 mn in the country by the year 2012, against a broadband target of 20 mn by the year 2010, and on equipment availability, favorable licensing condition and suitable business case, this definitely looks attractive.

Also, a lot of momentum is developing with VSNL proposing to roll out 2,000 WiMax base stations, covering 120 cities in two years, and BSNL moving aggressively from current presence in ten cities to thousand cities by end- 2007. The other companies that are aggressively following the lead are Bharti and Reliance Communications in addition to Dishnet. These developments are in tune with recent announcements by the government to provide broadband service in majority of villages in two years.

As far as WiMax equipment availability is concerned, the certified fixed equipment (802.16-d) is available from a number of international vendors with sound deployment experience. The same is expected for mobile equipment (802.16-e) for which interoperability tests have been carried out at three PlugFests in the last one year. Many of these products conform to Wave-II operability criterion and are in the process of certification. Sprint, a major user, has also set up an interoperability test facility at Arlington, Virginia (USA) to meet its rollout requirements as mentioned above.

The business model for WiMax (or WiBro) deployment depends on cost/availability of infrastructure associated with PTMP or fully mobile configuration. The latter includes tower, power, and backhaul capabilities. As far as CPEs are concerned, the fixed units (802.16-d) are more easily available and render longer coverage with outdoor directional antennae, but the Opex is high due to cost of individual alignments.

The other important issue is favorable licensing conditions including spectrum availability. While 50 MHz of license-free band has been identified in 5.8 GHz band, the government is still trying to have licensed allocation in 3.3–3.4 and 3.4–3.6 GHz bands. Some efforts are also in progress for spectrum allocation in 700 MHz band, which is well suited for rural application due to reduced propagation loss. Although low frequencies pose correlation problem in MIMO antennae, it is expected that new developments like 'Metamaterial Ultra-compact MIMO antenna array' from Rayspan may overcome the difficulty.

Based on the evolution plan of 3G technologies, there is also the possibility of sharing the FDD based 3G spectrum for FDD based WiBro technology, and TDD based 3G spectrum for TDD based WiMax. But, a word of caution: the above observation is purely based on current availability of FDD based WiBro and TDD based WiMax products. But, the two standards are quite flexible to use either of the two techniques and actual implementation would only depend upon industry preference based on intellectual property cost structure.

The author is executive director (R&D), Noida Software Technology Park