UHF RFID Leaps and Bounds...
From Systems to Silicon

Today's UHF RFID market is a somewhat typical, technology-driven, emerging market. The fundamental technology is being proven via limited pilots and trials of relatively small size. Important technology innovators are pairing with early adopters with enough buying power and influence to create markets on their own. While there is no lack of public and private capital investment, end-users of RFID must address the total cost of RFID ownership. This is not a simple ROI calculation but rather a complex financial analysis. Enter the chicken and the egg.

The chicken is the RFID reader, and the egg is the total available market for those readers (the corresponding TAM). Which comes first when they are so intertwined and dependent upon each other? Today's typical UHF reader is a four-port portal reader that sells in the market for thousands of dollars. It's big, heavy, consumes lots of power and provides very high performance for somewhat limited applications within the industrial market. If this complex reader were to be reduced down to an inexpensive chip set costing only hundreds or tens of dollars the TAM would explode. Reducing the system down to silicon can cause a paradigm shift within the market. Less capable but inexpensive chip set solutions will open up high-volume markets for deeply embedding UHF RFID capability into mobile and portable consumer applications not yet considered possible.

UHF readers are complex systems, comprising frequency synthesizers, filters, power supplies, RF transceivers, RF power amplifiers, interface communications and signal processing, packaging, memory and other items depending on the intended application. While each reader design is unique, all receive the same market pressure for increased value. Value is the utility of the reader versus its cost (value = utility / cost). Utility is derived from a mix of performance, size, reliability, functionality, system up-time and so on. Cost is actually the total cost of ownership and includes price paid, payment terms, maintenance/support costs and related items. Lower cost and higher utility result in higher value.

Additionally, lower utility with an offsetting lower cost will result in the same value. Today, UHF readers are high-value items with high utility and high costs. The RFID market demands high value at all cost levels, and a UHF reader chip set can provide the value at low cost levels capable of cracking open new and massive markets. Driving down cost and size will enable mass adoption and help pave the way to shelf readers, item level tagging, and a host of inexpensive mobile applications.

There are a number of challenges on the path to realizing a complete UHF reader chip set. First and foremost is possessing the requisite RF, semiconductor device and embedded software design expertise necessary to realize a high value design. The system architecture is a complex trade-off between performance, cost, process capability and product positioning.

For example Gallium Arsenide (GaAs) works well for RF functions but is poor for logic and purely digital functions. Complementary Metal Oxide Semiconductor (CMOS) is excellent for inexpensive digital logic circuits but struggles with high performance analog RF circuits. Silicon Germanium (SiGe) is a nice balance between the GaAs and CMOS extremes. SiGe is a low voltage process — +3V or +5V bias — that elegantly handles digital logic and RF power up to roughly +20dBm levels. GaAs, SiGe, CMOS and other process technologies can be leveraged and utilized for a chip set solution where the requisite digital and RF functions can be realized with separate chips; however, for SOC realizations the process choice becomes very complicated.

The second major challenge is that of design partitioning. From a top-level architecture standpoint most readers can be broken into four main functions: RF transmit amplification; RF receive signal conditioning; baseband processing; and host communications protocol. Specific choices of how to partition these functions can lead to many different physical implementations. One is a three-chip solution with a transceiver, a power amplifier and a baseband processor. The transceiver may include pulse shaping, frequency synthesis, modulation creation, automatic level control, up/down conversion and some RF power amplification.

A SiGe realization of this is WJ's WTX1000 RFID Transceiver Chip. A stand-alone, GaAs RF power amplifier capable of +30dBm transmit levels is optional. This approach keeps costs down for those not wanting the maximum allowed transmit power. For example, many printer applications work at low transmit power levels and do not need the power amplifier. Finally, the baseband processing function can be handled in any number of ways: microcontroller, embedded microprocessor, digital signal processor, Field-Programmable Gate Array (FPGA).

The last and critical item is that of host communications protocol choice and realization via embedded firmware. Regardless of how the chips are realized, a large amount of embedded firmware is needed to realize the RFID air protocol (reader to tag communications) and the communications protocol (reader command and control). It is important to consider the embedded firmware since it could affect migration from a chip set to a single, SOC solution.

Realizing a single, SOC solution is possible — today. With careful choice of semiconductor process one can realize the RF transceiver and baseband processing on a single die. All of the RF and digital design, semiconductor simulation tools, and automated layout capabilities exist, and there are a host of options for on- and off-shore fabrication, packaging and testing.

(click image for larger view)

Smaller, lower cost UHF RFID readers are coming of age. According to Venture Development Corporation (VDC) the $1.7B global RFID market in 2004 will grow to $5.9B in 2008. This is significant growth along a typical technology adoption curve. However, complete RFID chip sets and SOCs are just around the corner. The ability to realize an RFID reader chip set for on the order of $100 breaks the price elasticity barrier and fuels massive adoption. Its high value will cause a paradigm shift whereby many of the present barriers preventing widespread UHF RFID systems are eliminated. New consumer applications become possible, and the future of massive deployment becomes today's reality.

Richard L. Woodburn is responsible for product line management and business development of WJ Communications' RFID products. His current responsibilities include marketing, product definition, product development, branding strategies, channel management, marketing and communications, strategic/OEM relationships and operation P&L. Woodburn has held various technical, operations and leadership positions with fiber optics and semiconductor products. He earned a B.S.E.E. from California State Polytechnic University, Pomona and an M.B.A. from Santa Clara University.