Protocols
The front-end of an radio-frequency identification (RFID) system is composed of two components : readers and tags. Tags store unique identifications and are attached to objects; a reader performs the tag interrogation procedure to recognize an object by issuing wireless RF signals to interrogate the identification (ID) of the attached tag. Since tags are designed for an attempt of worldwide deployment in commercial or alike applications, they are supposed to be tiny, low cost and equipped with a simple circuit of limited computation and communication capabilities.
Most RFID tags are passive, they do not have on-tag power source and derive energy from the RF field generated by the reader to drive the circuit. When a tag and a reader are close enough, they can communicate with each other. For such a situation, we say that a tag is in the interrogation zone of the reader. Like other wireless communication systems, the RFID system also suffers from the signal interference problem. There are two types of signal interference. The first one is called the reader collision, a situation which occurs when multiple readers issue signals to same tags simultaneously. The other is called the tag collision, which occurs when multiple tags respond to a reader simultaneously. Collisions hinder and slow down the tag interrogation procedure. Therefore, reader anti-collision and tag anti-collision protocols are required to respectively reduce reader collisions and tag collisions to improve interrogation procedure performance.
Reader Anti-Collision Protocols
Some reader anti-collision protocols are proposed to reduce reader collisions based on the concepts of Time Division Multiple Access (TDMA), Frequency Division Multiple Access (FDMA) or Carrier Sense Multiple Access. TDMA-based reader anti-collision protocol divides the transmission time into intervals and a reader can only transmit messages in its assigned intervals. The assignment of intervals can be done in a distributed or a centralized way. Waldrop et al. propose two distributed TDMA-based reader anti-collision protocols, called Distributed Color Selection (DCS) and Colorwave.
A reader graph is first derived, where any two readers are defined to be adjacent and have an edge between them if they may have interference with each other. Each reader is assigned a colour which stands for a reservation of a specific time slot for transmitting signals. If all the adjacent readers are with different colours, the reader collision is avoided. In DCS protocol, the maximum number of colours (max_colors) is fixed, and a reader transmits only in its assigned color (time slot). On the contrary, Colorwave protocol has dynamic values of max_colors; it is a dynamic color assignment mechanism to minimize the required number of colours in the reader graph. With the reduction in the number of used colors, the efficiency of message transmission is increased.
FDMA-based protocols divide all available frequency bands into several non-interfering frequency channels. If a frequency channel is only assigned to a transmitter at a time, transmitters can transmit messages simultaneously without causing any interference. Ho et al. propose HiQ, which is a both TDMA-based and FDMA-based protocol. It attempts to minimize reader collisions by learning the collision patterns of the readers and by effectively assigning frequencies over time. HiQ depends on a distributed, hierarchical, and online learning scheme called Q-learning for determining frequency and time assignments. By interacting repeatedly with the system, Q-learning attempts to discover an optimum frequency over time. EPCglobal Gen 2 is a famous protocol that adopts FDMA technology to solve reader collision problem. Readers can choose separate transmission channels to avoid interference by the frequency hopping spread spectrum technique.
CSMA is another mechanism used to solve the reader collision problem. In CSMA mechanism, each reader needs to check before transmitting messages whether the carrier (the shared communication channel) is free or not. If the carrier is sensed to be idle, the reader sends out a message at once. Otherwise, the reader delays a random period of time and starts sensing carrier again. The European Telecommunications Standards Institute (ETSI) EN 302 208 Standard utilizes “Listen Before Talk” (LBT) mechanism that is based on the concept of CSMA to solve the reader collision problem.
Tag Anti-Collision Protocols
Several tag anti-collision protocols are proposed to reduce tag collisions. The can be categorized into 3 classes : ALOHA-based, tree-based, and counter-based protocols. The ALOHA, slotted ALOHA, and frame-slotted ALOHA protocols are ALOHA-based protocols. In ALOHA protocol, a reader first sends a command to make tags transmit their IDs. On receiving the reader’s interrogation signal, each tag in the interrogation zone independently waits for a random back-off time and then responds with its tag ID to the reader. If no collision occurs, during a tag’s ID response, its ID can be identified properly. In slotted ALOHA protocol, the random back-off time must be a multiple of pre-specified slot time. Frame slotted ALOHA protocol is similar to slotted ALOHA protocol except that the whole interrogation procedure is divided into a set of frames with each having a fixed number of time slots, and a tag can send its ID to the reader only in one randomly chosen slot during a frame period. ALOHA based protocol is simple, but has the tag starvation problem that a tag may never be identified properly for the reason that its responses always collide with others’.
The basic idea of tree-based protocols is to repeatedly split the tags encountering collisions into subgroups according to tag IDs until there is only one tag in a subgroup to be identified successfully. The protocols can be applied to tags with or without writeable memory. Tags with higher memory have higher cost. However, tags for such a kind of tags have better performance. The Query Tree (QT) protocol is applicable to tags without on-tag writable memory. In the protocol, the reader broadcasts a request bit string S with variable length to tags. A tag with an ID prefix matching S will respond its ID to the reader. When collisions occur, the reader broadcasts again with a longer bit string S0 or S1 to split colliding tags into two subgroups. The bit-by-bit binary tree is applicable for tags with writable memory. In the protocol, the reader broadcasts a request command first and each tag will respond with the first bit of its tag ID. If collisions occur, the reader will acknowledge the tags with 0 (or 1). Only the tag with the first bit being 0 (or 1) will respond with the next bit to the reader. In this way, the tags are continuously split into two groups. The other tree-based protocols, such as the EPCglobal Class 0, the tree-slotted ALOHA (TSA), Bi Slotted Query Tree Algorithm (BSQTA) and Bi Slotted Collision Tracking Tree Algorithm (BSCTTA) protocol, also utilize similar concept to split tags to solve the collision problem. The main drawback of tree-based protocols is that their performance is affected by the length of the distribution of tag IDs. In general, tree-based protocols has longer identification time latency than that of the ALOHA-based, but it does not have the tag starvation problem.
The concept of counter-based protocols is similar to that of the tree-based protocols. The major difference between these two kinds of protocols is that the former rely on static tag IDs for the splitting, and the latter rely on dynamically changing counters for the splitting. ISO/IEC 18000-6B, each tag has a counter initially set to 0. When a reader sends request to tags, every tag with counter value 0 can transmit its tag to the reader. When a collision occurs, the tags with counter value 0 randomly generate a random bit, 0 or 1, and add it to their counters. In this way, the tags with counter value 0 are split into two subgroups. Other counter-based protocols, such as Adaptive Binary Splitting (ABS) protocol, utilize similar concept to split tags encountering collisions. The counter-based protocols do not have the starvation problem. Furthermore, they have the stable property that their performance is not affected by the length of tag IDs or the distribution of tag IDs.
The table below shows a summary of the protocols for reader and tag collisions.
Collision Type | Category | Protocol |
Reader collision | TDMA | DCS |
Colorwave | ||
FDMA | HiQ | |
EPCglobal Gen 2 | ||
CSMA | ETSI 302 208 Standard | |
Tag collision | ALOHA-based | ALOHA |
Slotted ALOHA | ||
Frame Slotted ALOHA | ||
ISO/IEC 18000-6A | ||
Tree-based | QT | |
Bit-by-bit binary tree | ||
EPCglobal Class 0 | ||
TSA | ||
BSQTA | ||
BSCTTA | ||
AQS (Adaptive Query Splitting) | ||
Counter-based | ISO/IEC 18000-6B | |
ABS |
Air Interface Standard
The air interface standard primarily affects two components of the RFID system : the RFID reader (interrogator) and the tag by defining rules for communication between the two devices. In particular, an air interface standard specifies:
· Detailed protocol rules, including modulation and bit encoding so that communication rules that are relevant to the air interface can be distinguished from any other radio signal, including noise.
· The anti-collision algorithm that is used to differentiate an individual tag from others to maintain an open communication channel.
· Structures for commands and responses, including the specification of additional processes required when invoking commands.
· Some aspects of memory architecture, including the definition of specific memory areas for particular functions, size constraints on particular memories, whether and how this memory is locked.
There are two leading figures in the air interface standard, which are The ISO (International Standards Organization) and the EPC Global. The ISO has their 18000 standard and the EPC Global Center has introduced the EPC standard. EPC standard for air interface is not compatible with the ISO 18000 UHF (Part 6) standard. The EPC and ISO 18000 (Part6) standards has similarity in the aspect of dealing with the tracking of merchandise through the supply chain. The ISO 18000 (Part 6) standard only deals with air interface protocols, whereas the EPC standard also includes data structure.
There are several evolutions to the EPC standard, in which Class 1-Generation 1 is the current version of EPC. Unfortunately, it is not backward compatible with Class 0. Generation 2 was developed in the hopes to be backward compatible with Class 0, however, merging with the ISO 18000 standard will be difficult, if not impossible.
EPC Standards
EPC stands for Electronic Product Code, initially proposed by the Auto-ID Center as the next standard for identifying products. The objective is not to replace existing bar code standards, but rather to create a migration path for companies to move from established standards for bar codes to the new EPC. To encourage this move, the basic structures of the Global Trade Item Number (GTIN) was adopted, as GTIN is an umbrella group under which all existing bar codes fall. There is no guarantee that the world will adopt the EPC, but the proposal already has the backing of the Uniform Code Council and EAN International, which are the two main bodies that oversee international bar code standards.
The Auto-ID Center at MIT has been working towards the development of a standard specification for item level tagging in the consumer goods industry called the Electronic Product Code (EPC). This has led to the establishment of a new group called EPCglobal, a joint venture between the Uniform Code Council (UCC) and EAN International, which maintain the U.P.C./EAN bar code system among others. As stated in the name, the primary goal of EPCglobal is to make the final EPC standard an official global standard.
The current thrust of EPCglobal is known as UHF Generation 2 (UHF Gen 2), a “Write
Once Read Many” tag with more memory (96 bits vs. 64 bits) than preceding Class 0 and Class 1 tags. UHF Gen 2 will also provide a bridge to the eventual Class 2 High Memory full Read Write tag. Prior to UHF Gen 2, Class 0 and Class 1 were being utilized for EPC, but they were not interoperable. UHF Gen 2 will merge the Class 0 and Class 1 standards for a global, interoperable EPC standard.
EPC Tag Classes:
Currently, several classes of tags fall under the EPCglobal umbrella. The difference between Class 0 and Class 1 is in the data structure and operation. Class 0 tags are read only, while Class 1 tags are one-time writeable. The EPC standards specified 5 classes of tags over time.
The chip manufacturer can only program the Class 0 tag, whereas the Class 1 Version 1 tag can be programmed on the factory floor. While both classes are functionally equivalent under the EPCglobal classification system, Class 0 and Class 1 use different hardware technologies to implement the Identity tag functionality. Class 0 tags are programmed when they are manufactured (referred to as “Read-Only” or “R/O”), therefore assuring uniqueness of the tag ID. Class 1 tags can be programmed once, referred to as “Write Once, Read Many” or “WORM”, by the user, providing operational flexibility.
Notably, Class 0 and Class 1 tags also use different protocols, or “air interfaces” to communicate. The difference in protocol means that they cannot communicate with each other. Still, tags of both classes can coexist in an environment, but they require readers that “speak their language” to be identified.
EPC code structure
The EPC is a number made up of a header and three sets of data. The header identifies the EPC's version number, allowing for different lengths or types of EPC later on.
- The second part of the number identifies the EPC Manager, most likely the manufacturer of the product.
- The third, called object class, refers to the exact type of product, most often the Stock Keeping Unit (SKU).
- The fourth is the serial number unique to the item.
Example of EPC : 01.115A1D7.28A1E6.421CBA30A
ISO Standards
The International Organization for Standardization (ISO) is based in Geneva, and its standards is somehow equivalent to law in some countries. All ISO standards are required to be available for use around the world, so users of ISO RFID standards are assured with the conformance of their systems to the different regulations on frequencies and power output for each country where they do business. The ISO is very active in developing RFID standards for supply chain operations and their work is nearing completion on multiple standards to identify items and different types of logistics containers.
ISO Standards for Proximity Cards
ISO 14443 for "proximity" cards and ISO 15693 for "vicinity" cards both recommend 13.56 MHz as its carrier frequency. These standards feature a thinner card with higher memory space availability and also allow numerous cards in the field to be read almost simultaneously by using anti-collision, bit masking and time slot protocols.
• ISO 14443 proximity cards offer a maximum range of only a few inches. They are primarily employed in financial transactions such as automatic fare collection, bankcard activity and high security applications. These applications prefer a very limited range for security.
• ISO 15693 vicinity cards, or Smart Tags, offer a maximum usable ranginf from 28 inches from a single antenna to as much as 4 feet using multiple antenna elements and a high performance reader system.
ISO Standards for RFID Air interface
The ISO 18000 series is a set of proposed RFID specifications for item management that could be ratified as standards during 2004. The series includes different specifications that cover all popular frequencies, including 135KHz, 13.56 MHz, 860-930 MHz and 2.45 GHz.
• 18000 - Part 1: Generic Parameters for Air Interface Communication for Globally Accepted Frequencies
• 18000 - Part 2: Parameters for Air Interface Communications below 135 KHz
o ISO standard for Low Frequency
o ISO standard for Low Frequency
• 18000 - Part 3: Parameters for Air Interface Communications at 13.56 MHz
o ISO standard for High Frequency
o ISO standard for High Frequency
o Read \ Write capability
• 18000 - Part 4: Parameters for Air Interface Communications at 2.45 GHz
o ISO standard for Microwave Frequency
o Read \ Write capability
• 18000 - Part 5: Parameters for Air Interface Communications at 5.8 GHz
• 18000 - Part 6: Parameters for Air Interface Communications at 860 – 930 MHz
o ISO standard for UHF Frequency
o Read \ Write capability
o Read \ Write capability
o Targeted for same markets as EPC standards.
• 18000 – Part 7: Parameters for Air Interface Communications at 433.92 MHz
o Manifest tag for Department of Defense (DOD)
ISO Supply Chain Standards
These standards are used to identify different types of logistics containers and packaging, in addition to individual items.
• ISO 17358 - Application Requirements, including Hierarchical Data Mapping
• ISO 17363 - Freight Containers
• ISO 17364 - Returnable Transport Items
• ISO 17365 - Transport Units
• ISO 17366 - Product Packaging
• ISO 17367 - Product Tagging (Department of Defense)
• ISO 10374.2 - RFID Freight Container Identification
References :
[1] Zhang, Y., Yang, L., & T., Chen J. (Eds.). (2010). RFID And Sensor Networks : Architectures, Protocols, Security and Integrations. Singapore : CRC Press.
0 comments:
Post a Comment