Context

RFID technology is available today in a huge number of applications and it is very ambitious to claim addressing this subject exhaustively. Very detailed information on RFID technologies can be found in [1,2]. Thus, it is very important to define the scope of our work. Recently, RFID situations are evolving to pervasive context with no prior knowledge about the tags to be read (number or positions), neither their environment. As an example, Item-Level Tagging (ITL) is pushed by retailers in order to track individual item instead of whole pallet. The benefit is a better inventory precision in supply chain management, thereby leading to strongly reduced costs. Multiplication of tag number in uncontrolled locations and arrangements involves harder reading conditions where ad hoc systems are no longer possible (Fig. 1).

The following video presents a pervasive integrity checking application involving multiples RFID tags inside objects. This scenario is typical of the challenging conditions raised by reading RFID tags in uncontrolled environments.

Ubi-Check video

In this study, we focus our interest in pervasive context in UHF band. This choice is guided by several factors:

• Growing number of applications in UHF band in pervasive situations or uncontrolled scenarios (i.e. uncontrolled tag positions and/or environment),
• Performance improvement of read rate is desirable to win new markets over,
• Read reliability which must be improved to achieve trustable systems (strong need for link reliability),
• Application case: RFID portal in supply chain with Item Tag Level (ITL).

Pervasive situations in RFID: RF challenges

To understand the main RF problems in RFID, an illustration of link budget is available here: Link budget explanation. RFID link can suffer from several causes, and several solutions could be implemented in practice to mitigate them (Fig. 2).

The first identified issue is propagation fading. In a multipath environment, signal reflection and diffraction may combine destructively and locally cancel out the electromagnetic field. If the E-field amplitude close to the tag is too low to power up the IC, reading will fail.

The second one concerns tag mismatching due to mutual coupling. Indeed, if two or more tags are close to each other, mutual coupling increases and may lead to tag detuning. In this case, accepted power is considerably decreased and once again, IC tags will not be powered up.

The third case deals with tag self-shadowing or object shadowing. In uncontrolled environments, various types of objects could introduce such effects. If a tag is placed in such an area, local field will be very low preventing IC power supply.

At first sight, collision issue seems more a protocol issue than a physical one. But collisions become problematic when a lot of tags are powered-up at the same time. By illuminating reduced areas, collision problems can be mitigated.

To explain these sub-optimal performances, two major causes have been identified. The first one is tag coupling, which leads to a sensitivity loss of the tag chip. This issue is out of the scope of this study. The second one is due to multipath environment that causes large signal fading. Diversity techniques can be implemented advantageously to mitigate multipath propagation effects.

In such a context, two types of actions are considered: mechanical move (tag or reader antennas) or/and electronic reconfiguration of the antenna radiation pattern:

• Tag mechanical move: for example, rotation of tag volume or shake when possible.
• Reader antenna mechanical move: antenna can be mounted on a sliding structure; this allows displacing the reader antennas.
• Reader antenna radiation pattern: reconfiguration in directivity or pointing direction (beam steering).
• Reader antenna polarization: antenna polarization can be changed in order to change propagation condition.

Solutions without mechanical moves have industrial preference since their costs are reduced and no maintenance is needed.

Objective of “Pervasive RFID” study

The three main issues leading to reduced read rate and read reliability in pervasive situations are the following: anti-collision algorithms (tag level, and reader to reader), tag coupling, and field coverage. Illuminating the portal scene with suitable and reconfigurable radio coverage should improve read performance (velocity and reliability). Therefore, in this project, we propose new UHF RFID antenna systems, based on diversity techniques, in order to overcome fading problems. Two main constraints have been taking into account: to keep a low complexity level and a system size convenient with industrial needs. Numerical simulations have been used to develop the desired antenna systems. Moreover, unlike previous studies, we characterize the antenna performance in a real case environment. To our best knowledge, the contribution of antenna diversity has been demonstrated here for the first time in a real RFID application use case. To this end, a RFID test bed platform has then been defined, built and installed in our laboratory. This set-up – closed to industrials requirements [State of the art on UHF RFID portal systems and diversity techniques] – is a versatile tool extremely useful to conduct our experiments in real conditions and validate the proposed concepts. This new portal will be also one key element to propose specific studies to RFID industrials (for instance, a national project has been submitted in fall 2014).

Additional improvement for reading reliability can be achieved in specific context by using application-level information and strategies. In situations showing multiples tags, it is possible to leverage on tag multiplicity for data resiliency and integrity checking purposes. In particular, we can detect incorrect conditions at the application level, which can be used for both increasing read rate performance and performing read recovery when errors are detected.  Adaptive reading protocols could be proposed using these principles. Read recovery can be done following two very different strategies: “read-retry” style, where the reading infrastructure issues different attempts to read the missing tags with varying parameters. And data-recovery approach, where data for an object or a group or objects are spread intro multiple tags, using an error-resistant coding.

To summarize, our objectives are the following:

• To define, develop and characterize a new test bed platform to quantify improvement achieved by each proposed solutions.
• To propose and experiment adaptive reading protocols, capable of robust performance in adverse reading conditions (typical of pervasive environments)
• To design of a versatile UHF RFID antenna system with:
  • Spatial, pattern or polarization diversity
  • Low complexity driving system,
• To model numerically the proposed RFID portal to quantify the diversity add-value.
• To conduct a set of experimental campaigns with increased complexity, matching the typical situations of pervasive applications. In particular, the effects of overlapping tags (stacking), metal presence and water presence should be considered.