Think More Thunk

Summary of WiMAX World 2007

In spite of the recent events surrounding the shift of leadership at Sprint and the resulting uncertain future of Xohm (mobile WiMAX) services, the future for WiMAX remains strong. The WiMAX World show two weeks ago appeared to be the largest yet. There was less hype than at previous events and more real hardware on the show floor. Fujitsu Microelectronics had two demonstrations: one a setup between two laptops simultaneously demonstrating their baseband chip for 802.16d, and a second 802.16e demonstration between a gateway and a television with the Fujitsu components being in the client only. The 802.16e demonstration was not done wirelessly because of the interference on the show floor, but was still impressive. Atmel was showing their RF chips for both AT86RF535B and AT86RF525B MHz for 802.16d. These components are unique because they support IEEE 802.16-2004 (.16d) and 802.16-2005 (.16e) WiMAX applications.  Samsung displayed a broad array of 802.16e-enabled mobile devices; from phones to laptop computers. The phones had a combination of three wireless communication protocols running on them: Samsung's WiBRO/802.16e, wCDMA, and the T-DMB broadcast protocol for video content.

It appears that the hurdle standing in the way of Sprint achieving its optimistic WiMAX phone forecast is actually the power-hungry WiMAX chips. Many companies are now developing more linear power amplifiers that will increase the efficiency of WiMAX phones, however much of this work is still in the technology R&D stage. Freescale announced high-power LDMOS chips to aid power amplifier component development, and we expect to see more of these types of announcements over the next 6 months.

It is interesting that companies are implementing power amplifiers in the low data rate wireless world as well. However, these applications require power amplifiers to overcome the ambient noise and increase the range. For these protocols, like ZigBee and proprietary 802.15.4 implementations, power consumption is not the issue that it is for WiMAX, because WiMAX is an OFDM protocol that by default uses more gates for its FFT and thus uses more power to operate.

As the ZigBee ecosystem matured, we began to see companies like Tendril, Atalum, and Synapsense who are focused of node registration and management. This was a signal of the sector's maturity. Similarly we are seeing companies like Bridgewater Systems taking this role in the WiMAX world. Bridgewater actually manages the registration and subscription allowances for companies like Sprint, Verizon, and Bell Atlantic, as well as others. The company's architecture can accommodate mixed networks of 802.16e, 802.16d, CDMA, DSL, cable, and pretty much any communication protocol you can think of. For Bridgewater the addition of WiMAX to its portfolio of capabilities is simply a minor tweak to its protocol. We think that since Bridgewater handles Sprint's services today, they are likely to incorporate the Xohm service as well.

The lingering question of where the radios for 802.16e are remains. Those demonstrating their baseband chip capabilities were not forthcoming as to the source of their radio chips. We maintain that the market for non-WiBro, mobile WiMAX (IEEE 802.16e) is at best two and more likely four years away from really taking hold. The question will be whether more advanced emerging protocols, such as LTE, can be developed quickly enough to offer competition or whether the protocols will segment geographically.

Evolution of Wireless Connectivity Technologies

Communication today is more inclusive of our environment than ever before; in fact, the line between communication per sé and our daily activities and movements is no longer clearly definable, nor is commu­nication solely about direct human action and interaction.

Over the last year four shifts in the connectivity market emerged. These include (1) an apparent loss of Freescale in the UWB market, (2) a perhaps consequential shift for PulseLink, (3) a strong move by CSR towards integrating UWB into its offering, and (4) the consequences of releasing pre-standard 802.11n to the marketplace. These four shifts are the central theme of our latest wireless connectivity report.

Freescale, having purchased XtremeSpectrum, became the market leader in UWB with products based on the DS-UWB architecture. Freescale was the most significant competition to the Intel-led MB-OFDM UWB group that eventually dominated the WiMedia Alliance. However, Freescale dropped its UWB product line abruptly in 2006. Apparently Freescale made a corporate decision based on risk reduction with its UWB components due to a perceived inability to obtain a timely ROI. According to many in the industry, Freescale was in full production with its UWB chips and in the process of delivering those components to customers like Belkin when the corporate sector of Freescale decided to cut its losses, leaving several customers in the lurch. This opened the market for domination by MB-OFDM-based UWB methodologies.

While there are many companies and organizations with UWB product development strategies, the WiMedia Alliance and the USB I/F are in competition primarily with scattered and incompatible protocols. The primary target application for the WiMedia membership is a wireless bridge for USB, corresponding initially to home and consumer electronics markets. Secondary targets include enterprise and commercial venues.As the wireless connectivity protocols mature and gain market acceptance, the differentiation between Bluetooth, WiFi, and UWB begin to blur. The same players involved in promoting and developing UWB-based wireless connectivity options are deeply embedded in the furtherance of the IEEE 802.11 family of specifications that drive Wi-Fi into a competitive stance in the wireless broadband arena. While each protocol and technology under development today will eventually find its own unique application and niche within the market, the component OEMs are maintaining their stake in the various initiatives as a means to ensure a place in the wireless connectivity market. In fact, nearly every player in UWB and Wi-Fi has legacy participation in Bluetooth applications. Now it is simply a matter of letting the market and equipment manufacturers define the application of the various wireless connectivity options.

A Wireless Joke

After digging to a depth of 100 meters last year, Japanese scientists found traces of copper wire dating back 1000 years and came to the conclusion that their ancestors already had a telephone network one thousand years ago.

Not to be outdone in the weeks that followed, Chinese scientists dug 200 meters and headlines in the Chinese papers read: "Chinese scientists have found traces of 2000 year old optical fibers and have concluded that their ancestors already had advanced high-tech digital telephone 1000 years earlier than the Japanese."

One week later, the Greek newspapers reported thefollowing: "After digging as deep as 800 meters, Greek scientists have found absolutely nothing."

They have concluded that 3000 years ago, their ancestors were already using wireless technology.

Certified Wireless USB (UWB) Forecast Divergence

There was a recent article in the EETimes about Artimi and their WiMedia-compliant dual-mode wireless USB / Bluetooth chip (the A-150). According to the article, Artimi will enter production at the end of 2007 with an initial unit price of $15. By 2009 the price may drop to $5, making it really competitive by the standards of today's marketplace. Artimi is an interesting company with a unique market focus and great potential.

However, the article offers an unattributed market forecast for Certified Wireless USB chips in 2009 of 200 million units. It is unclear if the forecast is an annual shipment forecast or if it is a cumulative one; either way the number is more than an order of magnitude above the current WTRS forecast for these chipsets.

There are companies, such as Wisair and Icron, which do have customers that are shipping product into the marketplace, however in the case of Wisair it seems that the chips are not yet Certified Wireless USB - compliant components. Wisair is in the business of developing rf component IP, however Icron is today selling its cable-free USB products which perhaps will be someday be included as part of the wireless USB specification. This is reasonable given the inevitable pace of a standards group as it lays down foundational rules, defines applications, and creates testing and interoperability compliance measures. With the exception of some software applications there are no products or markets that achieved 200 million units in its second, or even third, year of growth.

Put in perspective, 200 million units shipped means one device for every man and woman in the United States between the ages of 15 and 64. That is an impossible adoption to gain in just a couple years. There may be 2 billion USB peripherals in the world, but that level of adoption took more than 12 years, and really did not become a large market until the computer manufacturers were shipping USB in every PC and Laptop.

It is valuable to consider how market forecasts fit into the broader environment before blindly reusing them. In this way one avoids the inflated expectations that come with hyping a market before its time, and also the challenges faced when product forecasts do not materialize in two years. It is much easier to think about ones forecast first, rather than updating that resume two years from now.

Comparison of forecasts recently released:

• Spring 2007 - EETimes/Artimi - "Certified Wireless USB chips to reach 200 million units in 2009"
• Winter 2006 - iSuppli/Focus Enhancements - "160 million units in 2009"
• Winter 2006 - WTRS - "18 million Certified Wireless USB chipsets to ship in 2009"
• Summer 2006 - ABI - "UWB shipments to reach 300 million in 2011"
• Spring 2006 - Instat/Harris Wiltshire & Grannis LLP - "13 million UWB products in 2006, 190 million UWB products in 2009"

Is the IEEE a critical factor in developing industry standards?

One of the largely unforeseen consequences of developing UWB transceiver products outside the IEEE specification process is that many of the performance measurements for future RF chips are not discussed in a neutral environment; this includes not only the companies who will manufacture and sell the chipsets, but also universities and government research labs who all discuss the pitfalls and issues that can arise. Once UWB development left the IEEE and moved into a purely corporate arena, pitfalls and weaknesses were not openly addressed and thus are unlikely to be given the opportunity to be fixed. It appears that the development of MBOFDM UWB at this stage is largely political, and not an open, technical endeavor.

Therein lies the difference between the WiMedia Alliance and the Certified Wireless USB I/F, and the RF components based on the IEEE 802.15.4 specification which have been adopted by the ZigBee Alliance. The 'ZigBee' component-level products on the market today are differentiated, even though they all adhere to both the IEEE 802.15.4 specification and the ZigBee standard. Depending on a given application, one particular vendor's chips are better suited than another's, due to a tuning of various performance metrics to match that application's requirements. The components that are beginning to emerge from the UWB-focused standards groups are all focused on the same applications and the various performance characteristics are largely unavailable to the public. This means that the UWB components are not yet being sold except to very early customers, but it also means that their performance is probably wanting.

Thus, not only is UWB still a pre-emergent wireless protocol but once it does enter into the marketplace in reality, it is very likely that other technologies such as WLAN and Bluetooth will have taken its place in the applications currently being targeted by many of the UWB component companies. At that point it will be simply a question of whether the force and impetus represented by Intel can carry UWB into the marketplace. For those companies who placed all their bets on MB-OFDM taking over the market, it may be time to look to new wireless horizons.

WTRS Analysis from CTIA Wireless 2007 Day 1: WiMAX, GPS, WiFi, and more

Today is the first day of CTIA Wireless 2007 in Orlando, Florida.

In general, the show is well attended. The big market players are talking about WiMAX, WiFi-enabled handsets, GPS advancements, and UltraMobile Broadband.

In the M2M Zone, most companies are displaying uses of cellular technologies in applications such as AMR, vending, traffic management, facility management, and point of sales terminals. Big name companies like Siemens are displaying a tremendous number of application segments across a range of geographic areas. The primary technologies used by Siemens in the M2M sector are WCDMA and GSM. The company has expressly stayed away from older CDMA, presumably for political reasons.

WiMAX is definitely generating a lot of noise, as evidenced by the completely full WiMAX session this afternoon with lines out into the hallway, however there are also rumors echoing the industry that Sprint may not have the market focus behind WiMAX that the press is leading people to believe. The verdict on the impact of WiMAX is still out, beyond the real use case of back-end connectivity. While companies are showing and talking about mobile WiMAX, it will be up to the carriers to actually make the market. Meanwhile alternatives are fast approaching. Until Mobile WiMAX can reliably travel beyond the radius of a given cell in a handoff to its adjacent counterpart, the use case is not all that viable.

WiFi-enabled handsets are at the point of moving into the market full stream. This advancement of the application is likely to be steady, and not gain market share as rapidly as some analysts believe. Very few products have gained 300 million unit 'Voice over Wi-Fi', or VoWLAN, adoption in the matter of a couple of years. Voice over WiFi requires a shift in the way people use products, and that takes time. Look at the adoption of the camera phone. That took a long time to reach mass adoption. While VoWLAN is likely to be quicker, given the number of new technologies that paved the way and softened consumer resistance and contributed to consumer education, it is still not likely to skyrocket.

That aside, it is encouraging to see handsets on the market globally, from BT, to Orange, to Verizon. With these key big participants, others are likely to quickly follow. This is further evidence that the WiFi Alliance has done a good job of developing interoperability and certification for these products, as well as looking to new market opportunities for WiFi that stretch the technology beyond the PC.

GPS now has the chance to become as pervasive as the camera phone in handsets. However, it will not mean adding yet another power-hungry chip to the handset. In the next couple years we are likely to see highly optimized GPS solutions from companies like CSR. CSR purchased two market leaders, NordNav Technologies AB and Cambridge Positioning Systems Ltd, earlier this year and already has a product available on the market which is a stepping stone to future developments. Key to CSR's strengths, the focus of adding GPS is on minimizing power consumption but more importantly on reducing the MIPS consumption of new features.

Among the less publicized no-shows at the conference, the only exhibitor to show up from STMicroelectronics was their sign, and that provided by CTIA show staff.

Finally, UltraMobile Broadband (UMB) is being promoted heavily as a new addition to the 3G family. More details will follow tomorrow. The fight between Qualcomm and the rest of the industry who is tired of their business model, remains a strong underlying motivation for many of the companies here at CTIA to develop new wireless alternatives.

Tomorrow, expect to hear more about WiMAX, MediaFlo, and UMB.

WTRS Analysis of CTIA Wireless 2007: Day 2

Today is the second day of CTIA Wireless 2007 in Orlando, Florida.

The general consensus from conversations we had today leads to the very real possibility that in fact Mobile WiMAX is not ready for product use, and current estimates for its availability are between 2009 and 2011. However, we are seeing a growing group of Fixed WiMAX applications emerging which involve active (and in some cases exceedingly profitable) service providers, component manufacturers, and end product OEMs.

Towerstream is a Fixed WiMAX service provider with installations in multiple metro areas throughout the US. The company excels at offering a nimble and efficient network. One interesting point to consider is that in many senses a Fixed WiMAX network enables quasi-mobile applications for its subscribers. Within a given Fixed WiMAX node area, it is possible for an individual using WiMAX to move around and remain connected. Also when you leave a cell and move into the coverage area of the next cell, the product simply reconnects after 1-2 seconds of lost signal. While this is not acceptable for high quality voice applications, it is not an issue for data applications. So in fact it may be that mobile applications of Fixed WiMAX will develop the use cases for Mobile WiMAX and in essence pave the way for user acceptance.

The idea that a suite of wireless flavors are important to enabling wireless broadband products is being echoed through the industry. Just as CSR has done, TI has integrated combinations of WiFi, Bluetooth, and FM into chips. TI was also showing a concept for displaying multimedia content on a phone using a DLP (digital light processor) projector. It looks like the final concept includes integrating the DLP projector right into the phone, which allows a person to display a movie or presentation right on the wall.

Another approach to wireless broadband is to use HSPA modules, built directly into laptops. Ericsson is strongly focused on HSPA technology, citing analyst projections of 100 million HSPA products shipped in 2010 versus shipments in the same year of 30 million mobile WiMAX widgets.

Tomorrow, expect to hear about WiMAX, MediaFlo, UMB, and UWB.

An Unexpected broadband market driver

Clearly VOIP is viewed as a strong driver of Broadband service and thus infrastructure in the U.S. market. An additional, and lesser understood, driver for broadband services and infrastructure development in the US is the enhanced capabilities of the Internet Protocol afforded by IPv6 (Internet Protocol version 6) which exponentially increases the number of available Internet addresses, enables the proliferation of enhanced mobile services and applications, and affords an overall increased security. ENUM, a mapping protocol in use by global directory clearinghouses, depends in large part on the deployment of IPv6 as this next generation of the internet enables the assignment of IP addresses to everything with connectivity to the internet. In this case, if you look up ENUM, Wikipedia will fail to correctly define the term; ENUM is not a simple telephone number mapping protocol. ENUM has the potential to expand its use beyond mapping of telephone numbers and email addresses to a universal and global assignment of individuals to their associated numbers (telephone, IP addresses, email addresses, residences, bank numbers, drivers license numbers, and passport numbers). This promises a true convergence of communications devices onto the Internet by facilitating the integration of telephone numbers and IP addresses. There are various reasons for the interest in developing a global ENUM network. Chief among these is a monitoring and control capability which will allow governments to administer the society they serve. It makes possible, among other things, a consumer ability to finish watching a television program once he gets home from a friend’s house. In fact the one year long ENUM trial in North America ended on February 17 of 2007, under the authority of the ITU. The results of the trial will be released to governmental bodies at the end of April 2007. Participating companies include GoDaddy, VeriSign, NeuStar (the clearinghouse which controls all IP addresses, landline phone numbers, and mobile phone numbers in North America), Qwest, InCharge Systems, Instra, DelTel, Evolving Systems, AG Design, Seiri, BellSouth, Encirca, and Verizon. The dark side of ENUM and IPv6 is its infringement on the fundamental rights afforded to Americans by their constitution. In some countries around the world, governmental observance of its populous is considered acceptable due to the benefits of security that result from losing individual privacy. However even technologists must consider the implications of enabling governmental tracking of its population as the misuse of this capability in the hands of even a transient 4-year administration may do much more harm than illegal and indiscriminant wiretapping.

A CMOS software defined radio that operates across 174MHz to 6GHz

If this software-defined radio works as advertised, IMEC has achieved quite a feat!

"The 130nm CMOS IC covers all standards from 174MHz to 6GHz. IMEC’s wideband software-defined radio (SDR) transceiver, named SCALDIO, is widely programmable to operate with all current and future cellular, WLAN, WPAN, broadcast and positioning standards in the frequency range between the 174MHz and 6GHz. The unique architecture of the multi-mode SDR transceiver has a power consumption and CMOS chip area comparable to current state-of-the-art single mode radios thereby fitting the performance, power and cost requirements for integration into next-generation mass volume mobile devices. Today, state-of-the-art multi-mode terminals have a limited flexibility of two or three standard modes and suffer from increased power consumption and bill-of-material cost due to lack of reuse of building blocks. IMEC realized world’s first functional silicon of a true SDR transceiver IC which can be widely programmed to operate with all present and future standards in the frequency range between the 174MHz and 6GHz. The SDR front-end is highly integrated to save cost and area. The reconfigurable radio front-end IC in 130nm technology uses only 1.2V supply voltage and has an active area of 7.7mm2. Depending on its configuration the SDR front-end has a power consumption ranging from 60 to 120mA which is comparable to a single-mode radio."

Link: About IMEC - Press Releases.

Using Piezoelectric Windmills to Power Wireless Sensor Networks

Four researchers from the University of Texas Arlington recently completed a study that uses a miniature windmill to convert wind energy into electric energy. This is surprisingly more efficient than the large scale versions and uses piezoelectric vanes which are called biomorphs attached to a crankshaft. Because piezoelectric material can generate high voltages with a minute magnitude of deflection, this design could be feasible for powering wireless sensor networks. The team recently demonstrated just that and wrote up the results in a paper: "Myers, Robert, Vickers, Mike, Kim, Hyeoungwoo, and Priya, Shashank. “Small scale windmill.” Applied Physics Letters 90, 054106 (2007)."