Adaptive Broadband ARTICLE in full:
(full text, no graphics)
-- start --
Adaptive Broadband
The Role of OFDM and its Variants in Broadband Wireless Access
Premise:
More than 40 companies, including Lucent, Nortel, ADC, Siemens, Hughes and Alcatel have built Fixed Broadband Wireless Access solutions. All are aware of Orthogonal Frequency Division Multiplexing (OFDM) and its variants, and only one company, WiLAN, has chosen to implement it in its product. Against this backdrop, a consortium of major players, who do not yet have a working product in the market, comes forward to propose an OFDM variant, Vector Orthogonal Frequency Division Multiplexing (VOFDM) as the standard modulation scheme for all solutions in this space. They claim that if VOFDM becomes the standard and is implemented on a broad scale ,it will enhance coverage and dramatically reduce cost. One member of the consortium holds 30 patents for the VOFDM coding scheme. Who is right and what does the answer suggest about the technology that will drive mass market success for wireless access solutions?
Adaptive Broadband proposes, and provides supporting data through this white paper, that VOFDM is not the best available modulation technique for fixed wireless broadband access. As the market evolves to portable and ultimately mobile offerings, OFDM variants are more likely to play a role. In fact, the timeline offered by the consortium for the availability of VOFDM would make it unlikely that it could play a role before then. With a point-to-point version first appearing at the end of 1999 and a point-to-multipoint offering not available until late in 2000, at best, this technology will arrive during the second phase of service deployment. The market will not wait that long as the opportunity to speed up access beckons loudly.
Finally, pairing VOFDM with DOCSIS to create a fixed wireless access solution is sub-optimal. Radio is a sufficiently different transport medium from coax and fiber that it requires a separate approach. Mobility only adds to the difference.
Where Does VOFDM Fit in a Fixed Broadband Wireless Access Product?
To answer the mass-market technology question, we first need to consider all the elements of a solution. Coding schemes like VOFDM, Time Division Multiple Access (TDMA) and others address how digital information is transmitted over the air. If we consider the "life" of a small element of user data (a packet) traversing a wireless access network, it becomes clear there are other issues to contend with. The packet must first bid for access to the wireless access network. Access is granted based on a service level definition and a system bandwidth allocation method. Once granted, the packet must be "modulated" so it can be carried by radio waves. It must then contend for bandwidth on the channel in the appropriate direction (upstream/downstream). Then it is sent, demodulated and handed off.
To accommodate packets moving through the system, a solution builder must decide on four key elements of a solution, i.e. how to:
ú grant bandwidth to users (Packet-on-Demand, channel-on-demand, or dedicated),
ú divide the radio channel between upstream and downstream traffic (TDD vs. FDD), and
ú modulate the data onto the radio carrier (a combination of VOFDM, TDMA, CDMA, etc. and QPSK, 16QAM, 64 QAM, etc.).
ú define service (Asynchronous (ATM) levels of service, proprietary bandwidth-on-demand, Data Over Cable Service Interface Specification (DOCSIS), etc.),
Clearly ,the choice of a coding/modulation scheme is but one of four key elements of a solution which must interact to achieve efficient, reliable transport in a variety of environments.
What Combination of These Four Key Technology Elements is Right for This Emerging Market?
The main objectives in building a solution are to provide the most service possible in a given bandwidth reliably, to cover the greatest possible geography reliably, to deliver consistent service and to drive the cost down to consumer levels.
Bandwidth Allocation One Packet at a Time
To achieve these objectives, the product must allocate its bandwidth to users only when they have data to send or receive. Adaptive Broadband's AB-Access product features an ideal solution for implementing this scheme. The AB-Access Media Access Control (MAC) Layer offers bandwidth to users a packet at a time. This Packet-on-Demand scheme ensures packets are not left empty unless no user data is present at that instant.
Handling Upstream and Downstream Traffic
The second element of bandwidth efficiency is dynamic support for the changing balance of upstream vs. downstream traffic. Web browsing is downstream intensive. Voice is symmetric with roughly equal up and downstream requirements. E-Commerce is upstream intensive. During a given day, the ratio of upstream to downstream traffic will shift dramatically as E-Commerce and voice during the day give way to web browsing at night.
Time Division Duplex (TDD) is the most efficient method for handling this ebb and
flow. AB-Access employs TDD, which groups all of the available bandwidth into a single radio channel. The channel is then switched rapidly from upstream to downstream as needed to handle the presented load. By contrast Frequency Division Duplex (FDD) divides the available bandwidth into dedicated upstream and downstream channels. The size of these channels is relatively fixed and not adaptable in real-time to the presented load. An inherently FDD system (such as DOCSIS 1.1) is constrained by this fixed upstream/downstream channel allocation and cannot efficiently support the variety of applications a TDD system can.
Coverage is defined in terms of cell radius and the ability to serve non Line-of-Site (LOS) locations. This is where VOFDM comes into play. The reason all but one of the current vendors with product in the market have chosen TDMA and other non-VOFDM coding schemes is that the first two generations of Broadband Wireless Access Solutions are fixed in place. They are not intended to move once installed.
Given a fixed location scenario, design engineers have very effective, significantly less-expensive options for handling multipath fade and non-LOS locations. By coupling TDMA with a very cost-effective dynamic equalization scheme, Adaptive Broadband has achieved exceptional coverage of LOS and near LOS locations with its AB-Access product. As the leader in sold Broadband Wireless
Access Solutions, Adaptive Broadband, relies on more than theory to demonstrate its point. Empirical data gathered from a study of urban and suburban areas, along with data from live customer installations, all provide strong support for the assertion that AB-Access has achieved effective coverage of LOS, near-LOS and first wall penetration scenarios.
Cost is the issue. Standardized mass market ASICs advantage all modulation schemes equally. That advantage is not unique to VOFDM. But, VOFDM is relatively more expensive than TDMA coupled with dynamic equalization. It takes significantly more processing power to implement due to the cycles required to reconstruct the original signal from the many instances sent. It reduces the transmit power by three to four decibels due to the many frequencies transmitted, which means smaller cell radii. Finally, it requires more expensive components, specifically linear Power Amplifiers (PAs). All of this without delivering a significant advantage in fixed wireless scenarios.
In fact, Nokia and Ericsson have developed working versions of an OFDM variant as part of their Magic Wand project, but have not introduced mass market product, based on it due to the cost considerations mentioned above.
Service Level Guarantees
The third element of a successful solution implementation is service level guarantees. Leading solutions like AB-Access have borrowed from the terrestrial data world and implemented ATM transport with its associated ability to define and guarantee a nearly infinite number of service combinations. In this way, products have the power of and are consistent with the ATM networks to which they are interfaced.
Single ASIC Cost Advantage
The fourth element is price. The new consortium forwarding VOFDM as a coding standard are simultaneously offering DOCSIS, from the cable modem environment, as the MAC Layer standard. In this way, VOFDM would specify coding and DOCSIS would detail bandwidth access and service definition. The logic behind such a suggestion is that the success of cable modems will result in DOCSIS ASICs that could provide a cost effective and mature MAC for wireless systems.
The fallacy is that cable modem DOCSIS will efficiently service wireless implementations, when, in fact, it will need to be modified for wireless to a point where it varies greatly from the cable modem variant. Eric K. Wilson of Phasecom Inc. addresses this issue in a recent Private Cable and Wireless Cable article where he notes that "[DOCSIS] accommodates hardwire cable but not wireless systems. Wireless modems require many changes and enhancements for successful system deployment. The related issues include frequency tolerance and tracking."
In the end, wireless transmission is significantly different from cable transmission and requires a unique MAC to achieve efficiency and reliability in a simple fixed environment. As Broadband Wireless Access solutions evolve to support portability and mobility, the uniqueness only grows. Once the case is made that DOCSIS is not a fit with wireless implementations, the economic argument goes away. This is borne out by the design decisions evident in the leading products, all of which have wireless MAC Layers like AB-Access.
Focusing back on the economics of the decision, the consortium proposal would at best result in a two ASIC design, one to support VOFDM coding and the associated modulation along with a second to host the DOCSIS MAC. By contrast, the wireless MAC solutions, like AB-Access, can, and shortly will, have single ASIC products hosting both the modulation and the MAC layer. Contrary to consortium claims, the cost advantage goes to the single ASIC design.
Why, Then, is OFDM Such a Hot Topic in Wireless Circles?
Variants of OFDM have an advantage in mobile scenarios. Dynamic equalization works by "training" itself on the multipath environment and setting up the appropriate adjustments. If the system location remains fixed, the multipath environment remains relatively consistent and this technique is very effective. In a mobile scenario, the interference scenario is constantly changing. In this case, OFDM variants excel.
In fact, Adaptive Broadband offers a Coded OFDM (COFDM) solution for Electronic News Gathering (ENG) in the broadcast TV market for just this reason. Vans, helicopters and fixed wing aircraft are able to achieve nearly flawless transmission while on the move through widely changing multi-path environments in the largest urban areas. It is exactly this ENG application and digital broadcast radio that are the first commercial applications for OFDM variants due to their mobile component. Other OFDM variants are targeted for G3 cellular/PCS implementations, again because the application is mobile.
We continuously evaluate OFDM variants for inclusion in the AB-Access solution but remain convinced that adaptive equalization is a much more cost effective solution for fixed applications. We are currently doing development work in collaboration with the University of Cambridge on an OFDM variant for possible implementation in our portable and mobile AB-Access offerings, which are targeted for release in the summer of 2000 and 2001 respectively.
Standardization of the Air Interface Is Fundamental to Mass Market Success
To achieve its mass market potential, the Broadband Wireless Access market must reach consensus on an air interface, much as the cellular/PCS industry has done. This allows multiple vendors to build the subscriber and network elements and have them coexist in a single system.
A significant part of the air interface is the modulation scheme. Adaptive Broadband believes the AB-Access design is optimal for fixed broadband wireless access and has been working to forward it as a draft air interface for the industry. It is simply a better solution for fixed environments.
The Service Provider Dilemma: Wait for VOFDM Standardization and Product Development or Seize the Market Today
There is really no dilemma at all. The demand for faster Internet and Corporate LAN access exists today with devices and networks running at megabits per second being throttled by bottlenecked access circuits running two orders of magnitude slower. Equipment like AB-Access is available today at speeds and prices that open up significant SOHO markets. New service providers are spending hundreds of millions to deploy networks now. Market share is being claimed.
Further, VOFDM is not the best available modulation technique for fixed wireless broadband access. As the market evolves to portable and ultimately mobile offerings, OFDM variants are more likely to play a role. In fact, the timeline offered by the consortium for the availability of VOFDM would make it unlikely that it could play a role before then. With a point-to-point version first appearing at the end of 1999 and a point-to-multipoint offering not available until late in 2000, at best, this technology will arrive during the second phase of service deployment. The market will not wait that long as the opportunity to speed up access beckons loudly.
Finally, pairing VOFDM with DOCSIS to create a fixed wireless access solution is sub-optimal. Radio is a sufficiently different transport medium from coax and fiber that it requires a separate approach. Mobility only adds to the difference.
-- end -- |
