IEC 62575-2: RFID for Item Management — UHF Air Interface Protocol

Overview of IEC 62575-2: RFID Air Interface Protocol

IEC 62575-2, published as part of the ISO/IEC 18000-63 series for item management RFID, defines the air interface protocol for passive RFID systems operating in the UHF frequency band from 860 MHz to 960 MHz. This standard, technically aligned with the EPCglobal UHF Class 1 Gen 2 specification, has become the dominant global standard for UHF RFID in supply chain management, retail inventory tracking, logistics, and asset management. The protocol operates on a reader-talks-first (RTF) principle, where the reader initiates all communication by transmitting a continuous-wave (CW) carrier that simultaneously powers passive tags and provides the downlink communication channel.

The standard supports multiple frequency bands within the global UHF range: 860-869 MHz in Europe (ETSI EN 302 208), 902-928 MHz in the Americas (FCC Part 15), 920-925 MHz in China, 916-923 MHz in Japan, and 920-926 MHz in Korea. The protocol is designed to operate reliably across these diverse regulatory domains through a flexible configuration of modulation parameters, data rates, and channelization schemes. Tags harvest energy from the reader’s CW carrier using a charge-pump rectifier circuit, store it in a reservoir capacitor, and communicate via backscatter modulation where the tag modulates its antenna impedance to encode data onto the reflected signal.

Physical Layer and Link Parameters

The downlink (reader-to-tag) uses DSB-ASK, SSB-ASK, or PR-ASK (Phase-Reversed ASK) modulation with Miller or FM0 baseband encoding. The reader data rate ranges from 26.7 kbps (lowest) to 128 kbps (highest), selectable via the Query command parameter. The Miller encoding family (M=2, 4, 8) provides subcarrier-based modulation that offers improved noise immunity and clock recovery at the cost of reduced data rate. FM0 encoding, on the other hand, provides the highest data rate (128 kbps) and is suitable for high-throughput applications where read reliability is less challenging. The choice between Miller and FM0 encoding represents a fundamental trade-off between read range and throughput. Miller-8 encoding achieves approximately 20-30% longer read range than FM0 under equivalent link conditions, owing to its superior noise rejection and more robust clock recovery, while FM0 offers roughly 4x higher uplink data throughput.

The PIE (Pulse Interval Encoding) scheme used in the downlink encodes data bits by varying the duration of the low-power pulse within each symbol period. A TReal (data-0) has a short pulse duration, while a TData-1 has a longer pulse. The reader defines the link timing parameters Tari (reference interval for data-0), RTcal (reader-to-tag calibration symbol), and TRcal (tag-to-reader calibration symbol) in the Query command. Proper calibration of these timing parameters is essential for link establishment. The standard specifies a Tari range of 6.25 us to 25 us, with the corresponding data-0 durations varying proportionally.

Tag Inventory and Anti-Collision Protocol

The inventory process uses a slotted random anti-collision protocol based on the Q-algorithm. The reader initiates an inventory round by broadcasting a Query command with parameter Q (0 to 15), defining a frame of 2^Q time slots. Each tag selects a random slot counter value between 0 and 2^Q-1 and decrements it during each subsequent QueryRep command. When a tag’s counter reaches zero, it responds with a 16-bit random number (RN16). The reader acknowledges the tag with an Ack command containing the matching RN16, after which the tag transmits its protocol control bits (PC) and electronic product code (EPC). The Q-algorithm dynamically adjusts Q based on observed collision rates: if no reply is detected in a slot, Q is decremented; if a collision occurs, Q is incremented. The optimal Q value minimizes the total inventory time and is typically in the range of 4-8 (16-256 slots) for most retail applications.

The standard also supports the Select command, which allows the reader to pre-select a subset of tags based on memory content criteria before initiating the inventory round. This filtering capability enables sophisticated operations such as singulating tags from a specific product category, location, or expiration date range. When combined with the Session and Inventoried flag mechanism, the protocol enables efficient inventory of very large tag populations by dividing them into manageable subpopulations.

Engineering Design Insights for RFID Systems

Antenna design is arguably the most critical factor in UHF RFID system performance. On the reader side, circularly polarized antennas with 6-9 dBi gain and an axial ratio below 3 dB provide the best trade-off between read range, polarization mismatch loss, and coverage pattern. For portal applications (dock doors, conveyor belts), dual-polarized antennas with polarization diversity can improve read reliability by up to 30% by mitigating the effects of tag orientation sensitivity. On the tag side, the antenna impedance must be conjugately matched to the chip’s complex impedance, which typically has a large capacitive reactance of -100 to -200 ohms at 915 MHz. Practical tag read ranges vary from 2-3 meters for small item-level tags (e.g., 20 x 20 mm for pharmaceutical bottles) to 10-15 meters for large pallet tags (e.g., 100 x 20 mm dipole designs).

Regulatory compliance presents another layer of complexity in global RFID deployments. The radiated power limits vary significantly by region: 4.0 W EIRP in the Americas (FCC), 2.0 W ERP (equivalent to 3.3 W EIRP) in Europe (ETSI), and 2.0 W EIRP in China. Engineers designing globally deployable RFID systems must ensure that the reader’s transmit power, frequency agility, listen-before-talk (LBT) or frequency-hopping spread spectrum (FHSS) implementation, and channel occupancy comply with the most restrictive regulatory requirements across target deployment regions.