Dobkin - The RF in RFID: UHF RFID in practice

To a quantum mechanic the whole universe is one godawful big interacting wavefunction but to the rest of us, its a world full of separate and distinguishable objects that hurt us when we kick them. At a few months of age, human children recognize objects, expect them to be permanent and move continuously, and display surprise when they arent or dont. We associate visual, tactile, and in some cases audible and olfactory sensations with identifiable physical things. Were hardwired to understand our environment as being composed of separable things with specific properties and locations. We understand the world in terms of what was where when. So one can forgive us for being disappointed that the computers and networks that form so large a part of our lives, and often seem so intelligent in other respects (at least on a good day), are clueless when it comes to perceiving and recognizing all these discrete physical objects we so easily detect and categorize. Why do we have to laboriously inform a computer database, by typing or mousing or tapping a screen, that a perfectly recognizable object has arrived at our doorstep? Why is so much human intervention needed for such a simple task?

It is to correct this deficiency of networked sensibilities that the field of automated identification (auto-ID) has arisen. Auto-ID includes any means of automating the task of identifying a physical object. To date, by far the most common means of doing so is to print a special machine-decipherable bar code on an object, and then image or scan the code using an optical transducer to extract an identifying number. One-dimensional bar codes (so named because information is obtained in traversing the pattern in a single direction, not because such patterns are in fact absent width and height) are easily deciphered and, in the form of the Universal Product Code (UPC) and its more modern descendents, nearly ubiquitous in the commercial world. Two-dimensional bar codes are also available, and pack more information into the same space. Optical character recognition (OCR) can be used to acquire information from conventional human-readable text, at the cost of an increase in computing requirements and decreased reliability. However, all optical methods of identifying an object have some deficiencies. Most fundamentally, the sensing device must be able to see the identifying mark: optical techniques require a clear line of sight. Not only objects, but dirt, paint, ink, and other objectionably opaque but relentlessly commonplace substances can distort or deface bar codes and other optical marks, obscuring the information optical auto-ID techniques require. Mechanical damage to the marks or labels degrades their readability. To store more data requires more space, or the use of finer markings visible from a shorter distance. Finally, data stored in printed marks on a surface is not readily modified or extended, save perhaps by wholesale replacement. While optical techniques for object identification are versatile and inexpensive, it is clear that in many cases another approach may be helpful.

To remedy some of the deficiencies of optical ID, we can turn to an alternative technique, radio-frequency identification (RFID). RFID is the use of radio communications to identify a physical object. RFID is really not one but a suite of identification technologies, because of the differing characteristics of the radio waves of varying frequency employed, and the differing approaches to operating the sensors that serve to identify individual objects. RFID has existed for more than half a century, but its widespread application has had to wait for inexpensive integrated circuits to enable small, low-cost transponders (the parts of the system that get attached to an object to be identified, more commonly known as RFID tags) to be fabricated. Over the last three decades, as the capability of integrated circuits has doubled and the cost per function halved about every two years, religiously attending to Gordon Moores famous law, new RFID applications have become economically feasible. In particular, since the mid-90s, a great deal of effort has been focused on the application of RFID in the manufacture and distribution of goods: supply chain management, where until recently the bar code reigned supreme. To serve the needs of manufacturing, distribution, and shipment functions, RFID tags must be very inexpensive, compact, mechanically robust, and readable from at least a meter or two away. As we will examine in more detail in , this combination of requirements has led to the choice of ultra-high-frequency (UHF) radio waves and passive RFID tags as the approach of choice for many supply chain applications, and it is UHF RFID technology that is the main topic of this book.