To achieve the best possible performance of your interrogation zone, you have to consider many aspects related to interrogation zone configuration, the number of tags in the zone, tag performance and velocity, external challenges such as environmental conditions as well as RF noise and interference, and many other issues. I will take you through some of these and together we will figure out a way to overcome these challenges and ensure that your RFID interrogation zone works.
To ensure that the tags are read by the interrogation zone, you have to consider the time that the tags are present in the zone. This is called dwell time, or time in beam. The dwell time depends on how fast the tags are moving through the interrogation zone, the data-transfer rate between the tags and the interrogator, and/or whether the tags are being read or written to. Generally, the longer the dwell time, the more successful the reading and writing operations will be. However, a long dwell time can be impractical, especially when you need to achieve a certain system performance and your conveyors are set for 300 to 600 feet per minute or your forklifts drive through the RFID portals at speeds around 5 kilometers per hour.
If you cannot change the speed of the tags traveling through the interrogation zone, you will have to change the zone parameters. You can increase the possible time in beam by configuring the antennas so that they will cover more space with their RF signal. You can do this either by positioning the antennas at an angle, so that the tag goes through the widest possible part of the beam (see Figure 2.1), or by sequencing the antenna transmissions following the direction of the tag's movement (see Figure 2.2).
Figure 2.1: Time in beam
Figure 2.2: Antenna sequencing
Interrogators use several types of antennas, and each type has slight performance differences. Antennas vary based on their polarization (linear and circular) and by their function (mono-static and bi-static).
Antennas can have either linear or circular polarization:
Circularly Polarized Antennas These produce a rotating RF field because they can receive signals from both vertical and horizontal planes and in between. These signals are a bit out of phase, which causes a certain loss of signal strength (50 percent on horizontal or vertical axes, less in between). However, because of the circulating field, often resembling a spiral, the antenna is not orientation sensitive and is suitable for use in applications where the tag orientation cannot be secured.
There are two types of circularly polarized antennas: right-hand circular and left-hand circular. With right-hand circular polarization, the rotation of the field is clockwise in the direction of the wave propagation. With left-hand circular polarization, the field rotates counterclockwise, again in the direction of the wave propagation.
You can implement right-hand and left-hand circular antennas in your interrogation zone to increase read efficiency.
Linearly Polarized Antennas In contrast to circularly polarized antennas, with linearly polarized antennas the RF beam is transmitted in one plane, which can be horizontal or vertical. A linear antenna produces a consistent signal in one direction with lack of phase distortion; therefore, it can have better penetration of certain objects and slightly higher efficiency when reading tags. These tags have to have the same polarization as the antenna and be properly oriented for successful communication. Linear antennas are very orientation sensitive; therefore, they should be used in applications where the tag orientation is consistent, especially when using single dipole tags. This is further discussed in Chapter 4, "Tags."
Antennas can vary by design; they can be bi-static or mono-static:
Mono-static Antenna A mono-static antenna fulfills both transmitting and receiving functions. An interrogator that supports mono-static antennas usually has a circulator that switches the transmitting and receiving functions and four ports intended for four antennas. An advantage of mono-static antennas is their smaller size as compared to bi-static antennas put into one case; however, they may be slightly less efficient due to using a circulator. Some manufacturers, however, such as Impinj, have tried to address this two-way communication challenge.
Bi-static Antenna A bi-static antenna uses two separate antennas: one for transmitting the RF signal to the tag and one for receiving the signal from the tag. These two antennas either can be integrated into one case, where each will have its own connector, or can be in separate cases. For the bi-static antennas to work, you always have to engage reciprocal antennas. Bi-static antennas are slightly more efficient than mono-static because they do not use a circulator and each antenna is solely dedicated to transmitting or receiving. An interrogator that supports bi-static antennas usually has eight antenna ports (four transmission ports, usually marked Tx, and four receiving ports, usually marked Rx) to accommodate four antennas.
Transmitting and receiving antennas have the same construction; their function depends on the port you plug them into. If you plug a bi-static antenna into port Tx, it will be used for signal transmission; if you plug it into Rx, it will be used for signal reception.
The antenna field consists of a main lobe, side lobes, and possibly back lobes. Now wait a second-back lobes? The RF signal can be diffracted (bent) to radiate to the back of the antenna-that's where the back lobes came from. The shape and size of all lobes depend on the antenna output power, antenna gain, and quality of antenna construction, as well as interference and reflections in the environment.
Most of the time, antennas do not have a perfect coverage within their lobes. These imperfections (sometimes called "holes") can be caused either by minor defects on the antenna reflector, by multi-path interference, or by external interference.
Multi-path interference is caused by reflections of an RF signal interfering with the antenna's field. When the reflected wave crosses the transmitted wave, it causes either null points with no signal or spots with a high signal concentration.
The antenna field is almost never perfect, and besides holes, its coverage usually varies in vertical as well as horizontal dimensions (see Figure 2.3). This is also usually caused by the antenna's construction and the unintentional imperfections in the antenna reflector as well as by outside interference or reflections. Before you install the antennas in your interrogation zone, you should test their function and make sure that you know their approximate coverage areas. You can do this with a tag on a yardstick or with a power tag, which measures the strength of an RF signal.
Figure 2.3: Perfect antenna coverage vs. reality
Do not forget that antenna coverage varies not only by the manufacturer, but also by each unit!
Antennas can be installed in many ways, but are usually grouped to achieve the best coverage of a given area. Most often, antennas are housed within portals or tunnels.
RFID portals are usually implemented at dock doors, conveyors, personnel doors, and various entrance and exit doors and gates. The skeleton of a portal is usually constructed out of a sturdy material such as steel, aluminum, or thick plastic. Antennas and reader(s) are attached to this construction. There are no rules defining how the RFID portal should look. It highly depends on the application as well as the portal and hardware manufacturer. Some companies provide RFID portals already assembled and ready to be put in. This portal or stand can come with specific hardware, which is usually the case with stands provided by the reader manufacturer, as opposed to portals developed by independent integrators, for which you can choose any hardware components that you would like to integrate.
RFID stands can come either open (you can see and reach all components) or closed (the components are enclosed within the stand). The enclosure is usually done by using RF-neutral materials such as plastic, which protects the reader and antennas from damage as well as from being tampered with by unauthorized personnel.
RFID portals can carry a number of antennas; they are limited only by their size and application. The dock door portals usually include two to four antennas on each side of the door. Sometimes these antennas are accompanied by one or two antennas mounted overhead for better coverage and penetration. Overhead antennas are used with side antennas for conveyor reading, dock door pallet reading, or with stretch-wrapping machines, for instance.
To modify the antenna field and change its direction and reach, it is useful to be able to rotate or move the antenna within the stand. If you are designing the stand or portal yourself, you can employ adjustable speaker holders used for home surround-sound systems or similar components to enable antenna positioning.
Usually, each RFID stand has a space for a reader. However, in many applications, you will need only one reader per interrogation zone. Therefore, all antennas at one dock door will connect to one reader that will be placed within one stand or in between the two stands. There are some restrictions regarding interrogator placement. You do not want to place the interrogator far away from certain antennas, because it would be necessary to use longer cables, which would cause a higher loss in signal strength. Also, the interrogator should stay out of harm's way, which is not always easy to accomplish, especially in a warehouse environment.
In some instances, you will use one interrogator for managing antennas belonging to neighboring dock doors. Now you are probably asking, "Why would anyone do that?" The main reason could be a problem with wiring the interrogator in order to reach antennas on both sides of the dock door. Many dock doors are sliding doors that open parallel to and closely spaced along the wall, which prevents you from installing the interrogator at the top of the door opening. Depending on the door mechanism, you may also have difficulty running the appropriate cables across the door. Using an interrogator to run antennas belonging to neighboring dock doors can provide the same results as the common "one interrogator, one dock door" approach. You will simply assign antennas to appropriate dock doors, and after the interrogator knows which antenna belongs to which dock door, it will be able to distinguish which tags were read where.
Another popular antenna configuration or a form of a portal is an RFID tunnel. Tunnels are used mainly with conveyor lines to focus the RF beam for successful reading of and writing to the tag. RFID tunnels are usually cubic enclosures made of sheet metal to prevent interference from the outside environment. On the inside, RFID tunnels contain antennas and RF-absorbent material (such as anechoic foam or ferrite) to absorb stray RF waves that could reflect and cause interference. A tunnel has front and back openings that are sometimes covered with a special curtain made of reflective material facing outside the tunnel and absorptive material facing inside the tunnel to prevent signal reflections.
To prevent reflections off the metal conveyor inside the tunnel, you can replace the existing metal rollers with rollers made of RF-neutral or absorptive material.
To achieve the best performance from your interrogation zones, you have to tune the antenna for proper coverage. You do not want your interrogation zones to overlap and interfere with each other, but you also want to achieve the best coverage in the area that the tags will most likely travel through.
Real World Scenario-Getting out of a Jam
At an RFID installation, I had to figure out how to read tags on cases full of peanut butter that were coming down the conveyor in various orientations. Side and overhead antennas could not penetrate the peanut butter to read any tags that happened to be on the bottom of the case. I figured out that there also should be an antenna pointing to the product from the bottom of the conveyor. However, because the conveyor used metal rollers, the bottom antenna may not do much good. Fortunately, there are rollers available that are made of a special plastic that is transparent for RF waves. They do not last as long as the metal ones, but you can afford to implement them in smaller areas where they are really needed. The only problem with plastic rollers can be the generation of static electricity (see Chapter 6), which has to be solved by proper grounding. Needless to say, the plastic rollers went together on this deployment like peanut butter and jelly.
Increased power input into the antenna will also increase antenna output, which will enlarge the antenna coverage patterns and increase the reading distance. You have to make sure that you do not cross the limits set by the regulatory agencies, however. Usually, interrogators manufactured in the country where you are using them comply with local regulations; therefore, you do not have to worry about the limits. This is, of course, valid only if you use standard equipment. If you have the equipment custom designed or if you use noncertified and noncomplying components, you must make sure that your final antenna outputs do not go over the specified limit or you could be fined for each day of using noncompliant equipment.
The signal strength emitted by the antenna does not depend only on the power input into the antenna, but also on the antenna gain and cable loss. A standard antenna manufactured in the United States has a gain of 6 decibels (dB), and cable losses count in the tenths of decibels. The readers are configured in such a way that they produce power that will provide for maximum allowed antenna output with the standard gain and supplied cables. If you use an antenna with higher gain or shorter cables, you will need to turn the power down at the interrogator level.
Don't forget that with changed antenna output, the location and size of holes in the RF field may be positively or negatively affected as well.
When you are setting up your interrogation zones and you know that some of the zones will overlap, you will have to shield each zone to prevent interference and reading of tags that belong to other zones. Do not forget that some other devices also may be affected by the RF emissions from your interrogators, such as RFID printers.
Materials most commonly used for shielding are metal mesh with holes significantly smaller than the wavelength of the frequency you are trying to shield, ferrite, or absorptive anechoic foam. You should not use metal sheets for shielding because the continuous metal surface could cause signal reflections, which could cause multi-path interference as well as interference at nearby interrogation zones.
The dense reader environment is defined as an environment where the number of simultaneously operating readers is larger than the number of available channels (for the United States it would be more than 50 readers in the area). You need to take into account several considerations that will affect your system.
Although dense reader mode is used to avoid reader-to-reader interference and increase efficiency in reading tags, it usually results in slower data-transfer rates between readers and tags. While in a single reader mode, the data rates can be up to 640 kilobits per second, but using a dense reader mode can slow them down about four times (or more).
Real World Scenario-I Don't Do Windows
In a distribution center, I had to implement an RFID printer that would be used for exception processing right next to the dock door RFID portal. After the printer was in place, it kept printing out void labels. I checked its configuration and everything else that could cause this problem but could not figure out what was wrong. I finally realized that the printer could not validate the RFID tags because of interference from the nearby interrogation zone. Although the printer had metal casing that provided shielding, it also had a plastic window on the side to view the status of the media. As soon as I covered this window with an adhesive metal mesh foil to keep the view but prevent the interference-problem solved.
A dense reader environment also affects network traffic. As with any other network device, the reader will use a certain portion of the bandwidth to send the data to the host. With dense interrogator installation, you have to make sure that the network bandwidth is not already fully exhausted by various devices (such as PCs, barcode scanners, handheld readers, and RFID printers). The effect on network traffic will also vary by the type of your readers. If your readers provide filtering and data aggregation, they will send a lot less data through the network than will readers without these processing options.
Operation in dense reader mode provides a little bit more sophisticated way to avoid interference between simultaneously working readers than using shielding. Dense reader mode uses several methods to achieve these goals, such as synchronization, listen-before-talk (LBT), and frequency hopping.
The transmitting and receiving functions of readers in multiple-reader environments have to be synchronized to avoid interference. Without synchronization, the signals from other interrogation zones can interrupt the tag-reader communication or cause misreads or issues with writing to a tag. Synchronization methods include multiplexing and software synchronization.
Two types of multiplexing are currently used with RFID readers, time-division multiplexing and frequency-division multiplexing:
Time-division multiplexing is based on readers sending signals on the same frequency in assigned time slots or operating for a certain time interval when other readers are turned off.
Frequency-division multiplexing is based on the frequency spectrum being divided into multiple channels, and each reader can have its own channel to operate on. Some frequency bands-such as the unlicensed UHF band in the United States, which ranges from 902 to 928 MHz and can be divided into 50 to 124 channels-by law require a reader to hop off one channel onto a different one after less than half a second. If there are more readers than number of available channels, the readers have to employ frequency hopping in conjunction with the listen-before-talk and Q algorithm approaches.
The listen-before-talk (LBT) approach is used with the frequency-division multiplexing scheme. The reader has to listen for any other reader transmitting on the chosen channel; only after it determines that the channel is available can it start using this channel for communication. If the channel is being used by another reader, the listening reader has to switch to another channel. This technique is mainly required in Europe by regulatory agencies.
Some interrogator manufacturers include a special LBT port for a single receiving antenna, which is used for listening while the reader is operating in the LBT mode.
Frequency hopping is a method of switching channels when operating in a dense reader mode. Readers may be required to hop across multiple channels within a given frequency spectrum, mainly because of the time interval restrictions for transmitting in one channel. In the United States, the UHF band (902–928 MHz) is divided into 52 channels (500 kHz each), which are then used for frequency hopping. Readers may be required to use LBT before crossing to another channel.
When two tags respond simultaneously to interrogation, the event is called a collision. If a collision occurs, the reader cannot successfully transfer data within either tag. To prevent tag collisions, readers and tags employ several anticollision mechanisms. The main anticollision methods are synchronous and asynchronous, also called deterministic and probabilistic, respectively.
The probabilistic, or asynchronous, method is based on tags responding at randomly generated times. This method includes several specific protocols. The most well known is the ALOHA protocol, developed by the University of Hawaii, which was originally intended to avoid data collision in early LANs. ALOHA mode is based on a node not transmitting and receiving data packets all at once, but instead switching these functions based on time. If a collision occurs, the node transmits the data packet after a random delay.
Generation 2 tags use a slotted ALOHA-based anticollision mechanism called the Q algorithm (so named because of the University of Hawaii networking guys who came up with the protocol to be used in wireline networking). The reader sends a query with a parameter Q and a session number to a tag, and then creates a slotted time. The tag generates a random 16-bit number as a handle. When the reader inventories tags in a selected session, the tag has to generate a random number for a slot number, which has to be between 0 and 2Q-1 . If this number is 0, the tag will send its handle to a reader. If the slot number is not 0, the tag waits a number of slots to send back the handle. If the reader acknowledges a single tag with a handle, it goes to an access phase and the rest of the tags have to wait for another round. If two or more tags answer, the reader has to send the same or a modified Q, and the whole process will run again.
The deterministic, or synchronous, reader protocol method is used by Generation 1 tags and is based on a reader going through the tags according to their unique ID. The most well-known synchronous method is a binary tree, or tree-walking, scheme. The reader has to know the tag IDs and then it searches the tree of all tag IDs. This method is slightly slower than the previous one because the search of all branches and sub-branches of the tree is time consuming, but the tags do not have to wait for the node to be available in order to communicate with a reader as well as hope for collision-free transmission.
Generation 2 supports two methods of backscatter encoding that tags use to send signals back to the reader. The reader assigns which method the tag should use when responding. These methods, Baseband-FM0 and Miller subcarrier, help with anticollision:
Baseband-FM0 encoding has been used by tags under the ISO standards but it is now supported by Generation 2 as well. This type of encoding is very fast but more susceptible to interference; therefore, it is not usually used in the dense reader mode of operation.
Miller subcarrier encoding is slower but less susceptible to interference because of an advanced filtering technique that helps separate the tag responses from the signal transmitted by the reader. This method fits the tag responses between the channels used by the readers. It also guards the readers to prevent them from crossing into the tag channels.