Software Architecture for a Multi-Protocol RFID Reader on Mobile Devices

Size: px

Start display at page:

Download "Software Architecture for a Multi-Protocol RFID Reader on Mobile Devices"

Trevor Allison
6 years ago
Views:

1 Software Architecture for a Multi-Protocol RFID Reader on Mobile Devices Joon Goo Lee Seok Joong Hwang Seon Wook Kim Sunshin Ahn Department of Electronics and Computer Engineering Korea University, Seoul, Korea seon@korea.ac.kr, KyungHo Park Ji Hoon Koo Woo Shik Kang Communication and Network Laboratory Samsung Advanced Institute of Technology, Korea Abstract In RFID (Radio Frequency IDentification) systems, a tag reader or a tag interrogator communicates with tags, reads their identification codes, and accesses their related database through a network infrastructure. There are many research activities in RFID hardware systems, but there is few in software infrastructure, especially on mobile devices. This paper presents software architecture and implementation of a multi-protocol RFID reader on mobile devices such as mobile phones and PDAs. We have designed and implemented most functionalities of a RFID reader in software except for code modulation and demodulation to support multi-protocols and the next coming standards easily. Also, in order to embed RFID reader s functionalities on WIPI (Wireless Internet Platform for Interoperability) based mobile devices, our firmware interacts with Handset Adaptation Layer (HAL). Any WIPI application can access our RFID reader system and query tags information from Internet through HAL interfaces. We have prototyped our system on the ARM-based Excalibur FPGA with ipaq PDA. Keyword: RFID, mobile, multi-protocol, WIPI, HAL 1. Introduction Currently we are thinking about assigning an unique identification and its associated information to an object for an ubiquitous environment. In order to support the environment, we need to develop a system to allow us to access objects information from objects themselves and from an existing network infrastructure such as Internet and sensor networks. This work was supported by Samsung Advanced Institute of Technology, Korea. There are many currently available solutions in data delivery and management on a network infrastructure. But there are still many constraints to capture data from versatile objects. One of currently existing solutions are manual data recording or manual scanning of bar-codes, and more advanced solutions are automatic scanning of bar-codes and sophisticated machine vision/sensing systems. All of these solutions are very expensive, time consuming, and inaccurate. Even they have more significant constraints in real work surroundings, for example, due to intensity of illumination and dirty spots on bar-codes. But in this case, we cost expensively even for getting a simple information. One of the potential and attractive solutions to capture data easily from objects is to use RFID (Radio Frequency IDentification) systems. In RFID systems, we attach a tag onto an object, and an identification code is stored in memory of the tag. A tag reader or a tag interrogator communicates with the tag, reads the identification code from the tag, and accesses its related database through a network infrastructure for getting more information. Figure 1 shows a typical environment for mobile-based RFID systems. In general, an industrial reader system in delivery and manufacturing services is implemented as many big and heavy RFID readers at fixed locations, and the system is highly optimized for accuracy and speed performance. The RFID system in a mobile environment has some different scenarios from an industrial reader system. First, because a reader can be carried anywhere and anytime to support an ubiquitous computing/networking environment, it is very difficult to implement a good system in terms of speed and accuracy. For example, when we use readers in a market place or in a museum, we will try to read tags for an explanation of goods or a work of art at walk. In this environment, a reader collision problem becomes severe. Many solutions have been proposed, such as coloring [1] and colorwave [2] methods. But these meth-

Figure 1. A typical example of mobile-based RFID systems. ods cannot be easily adapted to real mobile worlds due to limited performance of passive tags.

This makes a tag to be confused when multiple readers try to access a tag simultaneously.

2 Figure 1. A typical example of mobile-based RFID systems. ods cannot be easily adapted to real mobile worlds due to limited performance of passive tags. For example, passive tags cannot recognize more than one channels at one time. The tags only response at the same carrier frequency that a reader uses. This makes a tag to be confused when multiple readers try to access a tag simultaneously. Second, we must concern about flexibility of interface with mobile devices and RF modules more than performance of a reader. In a mobile environment a user doesn t want to read all tags around him/her. A user wants to read only a few tags at one time. In this paper, we focus on the second issue and present the software architecture for a multi-protocol RFID reader on mobile devices. There are many research activities in RFID hardware systems, but there is few in software infrastructure, especially on mobile devices. We have designed and implemented most functionalities of a RFID reader in software except for code modulation and demodulation to support multi-protocols and the next coming standards easily. The implemented software components are baseband modem APIs, a firmware including anti-collision engines, read/write functions and packetizer /unpacketizer of air interface commands/responses for each protocol. Also, in order to embed RFID reader s functionalities on mobile devices, our firmware interfaces with Handset Adaptation Layer (HAL) in the WIPI (Wireless Internet Platform for Interoperability) platform [3]. Through HAL interfaces, any WIPI application can access a RFID reader and query tags information from Internet. We have prototyped our system on the ARM-based Excalibur FPGA [4] with ipaq PDA. The paper is organized as the followings: Section 2 presents an architecture for a mobile multi-protocol RFID reader, and Section 3 describes the software architecture in detail. Section 4 presents implementation of our prototype system for verification, and Section 5 makes conclusion. 2 A Multi-Protocol RFID Reader for Mobile Devices In this section, we present an overview of our RFID reader architecture and discuss several major standard RFID protocols briefly. Figure 2 shows an overall software and hardware architecture of a multi-protocol RFID reader for our prototype system. The RFID reader consists of four parts: a RF circuit, baseband hardware logics, RFID software components, and a handset. The major modem hardware logic consists of a modulator and a demodulator. A modem controls a digital analog converter (DAC) for transmission, an analog digital converter (ADC) for reception, and a RF circuit for analog signals. While a modem is modulating, the transmitted data on a forward link are converted to rectangular pulses, and then filtered by a smoothing filter. In a double side band (DSB) transmission, a pulse shaped signal is transmitted to I and Q channels, which are tied to zero. In a single side band (SSB) transmission, I and Q channel signals are generated by a Hilbert transformer from the pulse shaped signal. These signals are mixed with a carrier frequency in a RF circuit and transmitted to tags finally. While a modem is demodulating, the received analog signals from

3 a RF circuit are passed to a bit extractor. These extracted bit symbols are decoded to logical bits, and then used in a firmware. The RFID software components consist of baseband modem APIs, a firmware, HAL interfaces, and WIPI-based applications. The baseband modem API provides interfaces to access baseband modem hardware logics and RF circuits for controlling modem hardware. The firmware includes anticollision algorithms, reader collision algorithms, interfaces with handset adaptation layer (HAL), and so on. The HAL is an abstract layer to provide hardware independence to an application [3]. Any WIPI-based application can access and control a RFID reader through HAL interfaces in an uniform manner. The software architecture will be described in detail in Section 3. # $ & ( * Table 1 shows forward and return link parameters and an anti-collision algorithm of each RFID protocol. It shows that ASK modulation is common to all the protocols, because it is easy to implement ASK modulation in a passive tag which has cost and power concerns. But there are many versatile baseband coding schemes [5]. Binary 1 and 0 can be represented in various line coding methods such as NRZ, Unipolar RZ, Manchester, Miller, FM0, and pulse time modulation (PTM) such as pulse width modulation (PWM) and pulse position modulation (PPM). RFID systems use one of these baseband coding schemes with considering an easy bit synchronization, a probability of error, a bandwidth, and receiver complexity. The details about coding schemes can be found in [6, 7, 8, 9]. An anti-collision algorithm of each RFID protocol is based on time division multiple access (TDMA) [5], because RFID systems are constrained by low computational ability, low power consumption, and little amount of memory. These constraints require a simple anti-collision algorithm. There are two major methods in implementation. One is ALOHA and the other is Binary Tree schemes. The original ALOHA is the simplest probabilistic method, but it is inefficient. For this reason, RFID systems use dynamic framed slotted ALOHA for efficiency. The Binary Tree scheme, a deterministic method of anti-collision, has many varieties such as Bit-by-Bit in EPC Class-0 and Bin slot in EPC Class-1 for fast identification rate. 3 Software Architecture Figure 3 shows an overall software architecture for a mobile-based multi-protocol RFID reader. We have implemented all RFID reader functions in software except for modulation and demodulation. It makes easy to support multiple protocols and to develop new standards in the near future. Our software architecture consists of four layers: baseband modem APIs, a firmware, a handset adaption layer, and a WIPI-based application [3]. Their details are described in the next subsections : - 9 > * & The software component for a RFID reader should allow users to control the hardware logics such as a modulator, a demodulator and a RF circuit, and it is provided in forms of baseband modem APIs. They provide interfaces to access baseband modem hardware logics and RF circuits for controlling modem hardware. Table 2 shows some example of our APIs. Additionally, the baseband modem APIs offer functions for transmitting /receiving bit streams from air interface commands and tag responses. There are various baseband coding schemes as shown in Table 1. If hardware does not support a certain encode/decode function, a software programmer can implement the function inside software manner in this layer or a transceiver part of each protocol. 2 $ > 4 The firmware includes three parts: an algorithmic part, a transceiver part, and miscellaneous functionalities. We separated an algorithmic and a transceiver parts for tag control because of easy programming and debugging. The algorithmic part includes an anti-collision engine, read/write and special function support. The transceiver part includes packetizing/unpacketizing functionalities for each protocol. Also the firmware provides HAL interfaces and miscellaneous functions. There are two types of anti-collision: a tag collision and a reader collision. Tag collision problems were solved by an anti-collision algorithm, as shown in Table 1. We considered air interface commands/responses and tag s state transitions for a tag collision arbitration. And we also meditated a reader collision problem. When multiple readers try to access one tag simultaneously, the tag may not answer because the tag accepts multiple channels. The tag may consider some or all of received signals as a noise. This implies that only one reader should access the tag at one time. In this situation, a coloring method using different channels may not work anymore [1, 2]. Another reader collision is occurred when a tag is replying to a correct reader and another reader attempts to send a command to the tag. In this case, the correct reader receives a tag response and an unexpected command together. If readers use different channels, the correct reader might extract the desiring tag signal from mixed signals. For these reader collision problem, we apply LBT (Listen Before Talk) and frequency hopping.

4 Figure 2. Overall software and hardware architecture of a multi-protocol RFID reader on mobile devices. Our firmware supports read/write functions and special commands. If a tag has user-writable memory, we need to provide read/write functionalities to access the memory. In this case, the firmware processes read/write commands with different parameters and methods for each supported protocol in one interface function. Also, the AutoID center, one of the famous RFID laboratory, proposes special commands, such as Kill command for protection of user privacy. EPC tags include this function essentially. If a tag is killed by this command with a password, the tag is deactivated forever. This helps you keep your privacy by stopping the tag s faculty before you get it. The firmware processes predefined commands with HAL. These commands have two types: a tag control and a reader control. The tag control commands include functionalities to read tag s ID and read/write data from/to user memory of tags. The reader control commands handle and examine the hardware including a RF circuit and a baseband hardware logic. The firmware packetizes and unpacketizes air interface commands/responses for each protocol. While a RFID reader is running, the firmware is waiting for a message from HAL. The received messages are decoded, and the firmware calls a proper function with the extracted parameters. We categorized HAL interface functions into tag associated functions, RF associated functions, reader associated functions, and reader environment associated functions. Each category and its example functions are shown in Table 3. There are many common functions to all protocols, such as CRC calculation of commands and responses, timer and delay, and so on. The functions are necessary to algorithmic and/or transceiver parts for most protocols, and they are in one functional block for efficiency and prevention of waste. 2 2 E G A RFID modem offers a firmware to applications through HAL, and HAL can access to modem logics and functions with this firmware. In WIPI, HAL is an abstract layer for hardware independence to be transplanted to any platform easily. The WIPI-based applications access hardware resources indirectly by calling HAL API functions. In this manner, any platform can be constructed independently regardless of hardware using the HAL API. The communication between HAL and the firmware is asynchronous. At first, HAL sends a command and a firmware accepts and excutes this command. After that, the firmware responses a message back to HAL. But in this manner, a program may wait forever until the response comes while channel is disconnected. Since this problem can occur, an asynchronous method is implemented in forms of Listeners in WIPI JAVA API and registered Callback functions in WIPI C API.

5 Figure 3. Overall software architecture of a RFID reader for mobile devices such as mobile phones and PDAs. Table 1. Link parameters and anti-collision algorithms in standard RFID protocols [6, 7, 8, 9]. Forward Link Return Link Protocol Modulation Baseband Modulation Baseband Anti-Collision coding coding ISO PIE FM0 Dynamic framed Type A slotted ALOHA ISO Manchester FM0 Binary Tree Type B ASK ASK EPC Class-0 PWM Subcarrier Binary Tree FSK (Bit-by-Bit) EPC Class-1 PWM BitCell Binary Tree Encoding (Bin slot) EPC Class-1 ASK or PIE ASK or FM0 or Dynamic framed Generation2 PR-ASK PSK Miller slotted ALOHA

6 Table 2. Example of baseband model APIs. Function name Description BBM PLL init Initialize PLL BBM TX ask modulation rate Set ASK modulation rate BBM TX mode Set TX mode to OOK or the others BBM RX gain analog Set RX gain BBM TX gain analog Set TX gain BBM TX link frequency Set forward link frequency BBM RX link frequency Set return link frequency BBM TX transmit bitstream Transmit baseband coded bit stream BBM RX receive bitstream Receive baseband coded bit stream BBM PA turn on/off Turn on/off PA BBM RF turn on/off Turn on/off RF Table 3. Examples of HAL interface functions in each category. Category Function Description Tag associated functions readids Read tags Identification codes writeuserdata Write data to user data memory readuserdata Read data from user data memory RF associated functions Reader functions associated Reader environment associated functions kill set/getrfstrength set/getregion setcrfr set/getreadcycles set/getregister getepc getmanufacturer getmodel Let a tag deactivate Set/get RF signal strength Set/get Region and its regulation Set carrier frequency Set/get reader s read cycles Set/get register value of hardware logics Get an electronic product code of reader Get a manufacturer of reader Get a model of reader

7 Figure 4. Our prototyped multi-protocol RFID reader for mobile devices. 2 H J & * & Wireless Internet Platform for Interoperability (WIPI) is the common platform envisioned by the Korean Ministry of Information and Communication for running mobile applications on any handset independent to vendors [3]. WIPI serves as a backbone for content providers to develop applications that run seamlessly on any mobile platform. WIPI supports for multiple programming languages such as Java and C, and downloads and runs content provider applications in a binary format from applications servers. Using WIPI, any WIPI application can access a RFID reader through Handset Adaptation Layer (HAL) and query tags from Internet for additional information with network APIs extended for RFID systems. 4 Implementation Figure 4 shows a picture of our prototype system. We implemented our baseband modem logic on the ARMbased Excalibur FPGA, ultra high frequency (UHF) RF circuit with filters, direct conversion up/down mixers, a PLL, a power amplifier and palm sized antennas. The RFID firmware is running on ARM9 processor on the Excaliber. And embodied baseband modem APIs control forward/return link frequencies, transmitting /receiving bit streams, PLL, TX/RX gains, and turn on/off each RF part. Our firmware has functions of inventory using anti-collision algorithms, and also has read/write functions if a protocol supports. Also special functionalities for security such as kill or lock at EPC protocols are accomplished. A HAL application was coded on a personal digital assistants (PDA) using a graphic user interface (GUI) based on PocketPC Using a universal asynchronous receiver/transmitter (UART), the HAL application on PDA and the RFID baseband modem were connected. When a user chooses a HAL command on GUI, the HAL application sends a message to the UART with predefined format and waits for a response from the firmware. The firmware of the baseband modem receives this message from the HAL application and decodes the message. After decoding the message, the firmware calls and executes a suitable function in an algorithmic part or a baseband modem API. If the firmware executes the function successfully, it sends back a response message, also a predefined format, to the HAL application. The HAL application displays proper information according to the responded message. 5 Conclusions and Future works In this paper, we presented software architecture and implementation of a multi-protocol RFID reader on mobile devices such as mobile phones and PDAs. There are many research activities in RFID hardware systems, but there is few in software infrastructure, especially on mobile devices. We have implemented most functionalities of a RFID reader in software except for code modulation and demodulation to support multi-protocols and the next coming standards eas-

8 ily. Our software architecture consists of four layers, such as baseband modem APIs, a firmware, a handset adaption layer, and a WIPI-based application. The baseband modem APIs allow us to control hardware logics such as a modulator, a demodulator and a RF circuit, and the firmware includes RFID protocol engines. Also, in order to embed RFID reader s functionalities on mobile devices, our firmware interfaces with HAL on the WIPI platform. Through HAL interfaces, any WIPI application can access a RFID reader and query tags information from Internet. We have prototyped our system on the ARM-based Excalibur FPGA with ipaq PDA for verification. Our on-going project is to make a baseband modem chip including all hardware logics, such as a modulator and a demodulator, and software parts such as a firmware and baseband modem APIs. This work is almost over. And we have a plan to develop a programmable RFID processor which eliminates all dedicated hardware logics for low power consumption, small size, efficiency and flexibility. References [1] Daniel W. Engels and Sanjay E. Sarma. The reader collision problem. IEEE Transactions on Systems, Mans, and Cybernetics, 3, October [2] James Waldrop, Daniel W. Engels, and Sanjay E. Sarma. Colorwave: An anticollision algorithm for the reader collision problem. In International Conference on Communication, volume 2, pages , May [3] Wireless Internet Platform for Interoperability (WIPI). [4] Excalibur devices. [5] Klaus Finkenzeller; translated by Rachel Waddington. RFID Handbook. John Wiley & Sons, John Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ, England, [6] ISO/IEC FDIS , ISO standard document. [7] EPC class 0, EPC global standard documents. technology/specification.html. [8] EPC class 1, EPC global standard documents. technology/specification.html. [9] EPC class 1 gen 2, EPC global standard documents. technology/specification.html.

Dynamic Framed Slotted ALOHA Algorithms using Fast Tag Estimation Method for RFID System

Dynamic Framed Slotted ALOHA Algorithms using Fast Tag Estimation Method for RFID System Dynamic Framed Slotted AOHA Algorithms using Fast Tag Estimation Method for RFID System Jae-Ryong Cha School of Electrical and Computer Engineering Ajou Univ., Suwon, Korea builder@ajou.ac.kr Jae-Hyun