32nd Annual International Conference of the IEEE EMBS Buenos Aires, Argentina, August 31 - September 4, 2010
Development of a Zigbee Platform for Bioinstrumentation Carlos A. Cifuentes, Gabriel G. Gentiletti, Marco J. Suarez and Luis E. Rodríguez. Members, IEEE
Abstract— This paper presents the development of a network platform which allows connecting multiple individual wireless devices for transmitting bioelectrics and biomechanics signals for application in a hospital network, or continuous monitoring in a patient’s diary life. The Zigbee platform development proposal was made in three stages: 1) Hardware development, including the construction of a prototype network node and the integration of sensors, (2) Evaluation, in order to define the specifications of each node and scope of communication and (3) The Zigbee Network Implementation for bioinstrumentation based on ZigBee Health Care public application profile (ZHC). Finally, this work presents the experimental results based on measurements of Lost Packets and LQI (Link Quality Indicator), and the Zigbee Platform configuration for Bioinstrumentation in operation.
I. INTRODUCTION
B
IOLOGICAL and physiological parameters as blood gases, blood pressure, heart beats per minute, respiration, temperature, electrocardiography (ECG) and others, must be measured and monitored in clinical diagnostic situations and monitoring of homecare patient treatments. In most cases, these measures are carried out through direct cable connections from the sensors to the equipment, causing a reduction in the flexibility because they are connected to the patient and the monitoring equipment. Introducing wireless networking systems for reading data from biomedical sensors allows great flexibility for both the patient and the medical staff. The use of wireless networking technologies ensures a digital management of data, thereby it improves the management of medical history. On the other side, the data can be stored for later analysis and can be consulted at any time from the hospital unit, or remotely through computers or PDAs, allowing the medical staff to monitor patients without going into their locations [1]. Integration of health monitoring telemedicine systems is an emerging area in information technology, which will enable early detection of abnormal conditions providing prevention of serious consequences. Continuous monitoring into the diagnostic procedure may benefit many patients who are under observation for chronic conditions or monitoring Manuscript received June 25, 2010. This work was supported in part by ECCI University, UNER Engineering Faculty, and ECI University. C. Cifuentes is with the Engineering Faculty Program at the ECCI University, Bogotá Colombia (e-mail:
[email protected]). G. Gentiletti is with the Engineering Faculty at the UNER University, Paraná Argentina (e-mail:
[email protected]). M. Suarez is with the Engineering Faculty at the ECCI University, Bogotá Colombia (e-mail:
[email protected]). L. Rodriguez is with the Electronic Engineering Faculty at the ECI University, Bogotá Colombia (e-mail:
[email protected]).
978-1-4244-4124-2/10/$25.00 ©2010 IEEE
recovery from any acute event up to surgical procedure [2]. A device for monitoring physiological parameters of type wearable on wireless biomedical sensors can be integrated into the user’s clothing [3]. Technological advances in instrumentation oriented to wireless networks [4], along with manufacturing miniaturized [5] and the integration of biomedical sensors with microcontrollers and radio frequency interfaces embedded in a single chip [6], promise a new generation of systems able to record wireless biomedical sensors and ready for the development of multiple medical applications. II. BACKGROUND IEEE 802.15.4 is a standard that defines the physical (PHY) and MAC layers for networks with low rates of data transmission, for use on portable devices. Zigbee technology defines the network, security and application framework for an IEEE 802.15.4-based system. These capabilities enable a network to have thousands of devices on a single wireless network. The ZigBee Alliance’s focus on the healthcare space has resulted in the development of the ZigBee Health Care public application profile (ZHC). It was designed to be used by assistive devices operating in non-invasive health care. ZHC provides an industry-wide standard for exchanging data between a variety of medical and nonmedical devices [7]. The principal application domains and use cases catered for are: Disease Management, Personal Fitness Monitoring and Personal Wellness Monitoring [8]. Nowadays, there are a variety of network nodes for Zigbee applications ready for data transmission, for example, Digi International Inc. offers XBee modules with board dimensions of 24.38 x 27.61 mm. and communication range up to 90 meters with line of sight, and XBee Pro modules with board dimensions of 22.0 x 32.94 mm with higher transmission power and communication range up to 1600 meters with line of sight [9]. This work aims at developing a ZHC network based on wearable biomedical sensor nodes, to be used in a hospital network, or in continuous monitoring in patient’s daily life as a body sensor network BSN, trying to affect minimally the patient’s comfort, for such reason, this device must have the smallest size possible and lower power consumption. However, to integrate a network nodes for Zigbee applications ready for data transmission, imply the addition of a microprocessor and complementary circuits to do functions such as acquisition or processing a biomedical signal, generating PCB size and cost increasing, and higher power consumption.
390
III. MATERIALS AND METHODS The Zigbee platform development proposal was made in three stages: (A) Design of hardware, including the construction of a prototype network node and the integration of sensors, (B) Evaluation of the system, definition of specifications of each node and scope of communication. (C) Implementation of a Zigbee network base on ZHC profile. A. Hardware Development The network node developed is based on Zigbee chip MC13213 composed of one 8-bits HCS08 microcontroller (9S08GT60) and a ZigBee transceiver (MC13192). Additionally, it is necessary to integrate some components that allow for wireless communication. We used a singletype circuit for antenna matching, which uses a single balun for transmission and reception, and switching of these functions is done internally, this topology has a reduction of components, but there is a greater sensitivity to noise [10]. It uses a chip antenna type (Fig. 2a) with a single matching for a small hardware footprint. The final prototype is shown in Figure 2b. We implemented a sensors network node module to evaluate the performance (Fig.2c). This includes connectors to biomedical signal acquisition modules, input/output peripherals, and acceleration, pressure, and temperature sensors built into the board.
each received packet and the LQI to the computer. At the end of the test, the lost packets were totalized, and The LQI was averaged.
Fig. 3. Experiment Setup for Evaluation.
There were three experimental sequences based on different payload sizes of transmit data packets, thus: 5 bytes, 10 bytes and 20 bytes, were used to represent the biomedical signals, and each sequence measurements were done at the following different powers transmission: low -8.0 dBm (0.154 mW), half 0.0 dBm (1.0 mW) and high 3.6 dBm (2.29 mW), and the effect of distance between devices was evaluated from 0 m and 100 m, with steps of 10 m each. Each measurement was repeated 4 cycles at different times, seeking to maintain line of sight between transmitter and receiver. The measurements were made in two different locations in an outdoor parking and an underground parking, in the last site there was not cellular signal or Wi-Fi networks. These locations were in constant activity and many cars and people were coming in and out constantly; in order to simulate the environment of the patient. C. Zigbee Network Implementation We were based on the tests developed and the Freescale Zigbee stack for implementing a bioinstrumentation network based on ZHC; this is a star network configuration where a ZC (Zigbee coordinator) receives patient’s signals data from ZED (Zigbee end devices). IV. RESULTS
Fig. 2. (a) Antena Chip 2.4 GHz Johanson Technology [11]. (b) IEEE 802.15.4 device developed. (c) Sensor Network Node Module.
B. Evaluation The evaluation of the modules was performed by an embedded firmware in the microcontroller. This was developed with the MAC layer on the Zigbee stack. The experiments were performed to assess the quality of biomedical signal transmission into patient ambient, based on data packets continuous transmission, where performance was evaluated on parameters such as: power transmit (from 16.6 dBm to 3.6dBm); distance between devices; communication channel (16 in 2.4 GHz) and payload packet size. We used two sensor network node modules for this experiment (Fig. 3): to start the test. One device had a transmitter profile, and another as a receiver. The transmitter sends 1000 continuous packets to the receiver, this transmits
The experimental results were based on the lost packets with values between 0 to 1000 and LQI within typical values between -95dBm to -18dBm according to IEEE 802.15.4 [12]. Below are the results of the experimental measures of three sequences, with 5, 10 and 20 bytes of payload size per packet, and a Zigbee Network implementation with biomedical signals. In figures 4 and 5 we can see the results that correspond to the test of 5 bytes in payload per packet. With low power transmissions we obtained a data reception up to a distance of 20 m in both scenarios (underground and outdoor). With half power transmissions, up to 50 m; with high power transmissions, up to 60 m, which was the maximum we obtained. At greater distances all packets were lost. Figures 6 and 7 show LQI measures and lost packets respectively for 10 bytes in payload per packet, with low power we got a measure up to 20 m in both scenarios; with half power, up to 50 m; and high power over 100 m.
391
Figures 8 and 9 show LQI measures and lost packets respectively for 20 bytes in payload per packet, with low power we obtained measure up to 30 m in both scenarios; with half power, up to 80 m; and high power over 100 m. The Zigbee network was put into operation (Fig. 10a), and we implemented a PC software for displaying the integrated sensor signals such as pressure, temperature, accelerometry, ECG, and indicators of LQI and lost packets (Fig. 10b).
Data Transmissions - 10 bytes in payload per packet 1000 Low Power Outdoor Half Power Outdoor High Power Outdoor Low Power Underground Half Power Underground High Power Underground
900 800
Lost Packets
700 600 500 400 300 200
Data transmissions - 5 bytes in payload per packet 100
-20
Low Power Outdoor Half Power Outdoor High Power Outdoor Low Power Underground Half Power Underground High Power Underground
-30
LQI (dBm)
-40
0
0
10
20
30
40 50 60 Distance (m)
70
80
90
100
Fig. 7. Graphics of lost packets vs. distance between devices at different power levels with 10 bytes of data.
-50
Data Transmissions - 20 bytes in payload per packet
-60 -20
Low Power Outdoor Half Power Outdoor High Power Outdoor Low Power Underground Half Power Underground High Power Underground
-70 -30
-80 -40
0
10
20
30
40 50 60 Distance (m)
70
80
90
LQI (dBm)
-90 100
Fig. 4. Graphics of LQI vs. distance between devices at different power levels with 5 bytes of data.
-50 -60 -70
-80
Data Transmissions - 5 bytes in payload per packet 1000 Low Power Outdoor Half Power Outdoor High Power Outdoor Low Power Underground Half Power Underground High Power Underground
900 800 700
-90 0
500
30
40 50 60 Distance (m)
70
80
90
100
Data Transmissions - 20 bytes in payload per packet 1000
400
Low Power Outdoor Half Power Outdoor High Power Outdoor Low Power Underground Half Power Underground High Power Underground
900
300
800
200
700
0
10
20
30
40 50 60 Distance (m)
70
80
90
Lost Packets
100 0
20
Fig. 8. Graphics of LQI vs. distance between devices at different power levels with 20 bytes of data.
600
Lost Packets
10
100
Fig. 5. Graphics of lost packets vs. distance between devices at different power levels with 5 bytes of data.
600 500 400 300 200
Data Transmissions - 10 bytes in payload per packet 100
-20
Low Power Outdoor Half Power Outdoor High Power Outdoor Low Power Underground Half Power Underground High Power Underground
-30
LQI (dBm)
-40
0
0
10
20
30
40 50 60 Distance (m)
70
80
90
100
Fig. 9. Graphics lost packets vs. distance between devices at different power levels with 20 bytes of data.
-50 -60 -70
-80 -90 0
10
20
30
40 50 60 Distance (m)
70
80
90
100
Fig. 6. Graphics LQI vs. distance between devices at different power levels with 10 bytes of data.
Fig. 10. (a) Zigbee Network proposed for bioinstrumentation. (b) Network management and visualization of biomedical signals software.
392
Finally, in figure 12, an application developed named Physical activity monitor PAM (Device ID 0x102b) by ZHC is shown, this is for kinematic data measures (typically acceleration, speed, position, and orientation) of a point in space. This application belongs to the category Health and Fitness Devices [8]. We measured the joints acceleration of ankle, knee and hip, this experiment was for five steps, in order to model gait patterns, figure 12a shows the three PAM ZED configuration with acquisition of three-axis acceleration and continuous transmission every 70 mS to a node ZC. Figure 12b shows the y-axis acceleration for each joint obtained by bioinstrumentation platform. The distance from the ZC to the PAM ZED did not exceed 10 m during the test, Figure 12c shows PAM ZED LQI obtained by ZC, it was not less than -40dB and the packet lost indicator was zero in this experiment.
In this evaluation there was not used retransmission of lost data, in order to analyze the sensitivity of the evaluated parameters and selecting the optimal values with fewer losses. With this parameters selected, we implemented retransmissions for lost data packet in the Zigbee network developed, it was for ensuring integrity of biomedical data and confiability in the communications. Finally, the implementation of Zigbee network based on a network star topology has the possibility to be a tree topology with the integration of ZigBee routers. Having the Zigbee platform for biomedical instrumentation in operation, we are working on developing a biomechanical analysis and rehabilitation system based on accelerometers and the integration of additional sensors. Besides, this application, based on this platform and web services applications, we are going to build web knowledge databases for patients health status monitoring and get reports through real time. For this new stage, we are planning to make evaluation with real patients in Teleton University Hospital. REFERENCES [1]
Fig. 12. (a) Three PAM ZED configuration. (b) Acceleration in the yaxis for Hip, Knee and Ankle (c) PAM ZED LQI measurements.
V. CONCLUSIONS With the conditions of this experiment, IEEE 802.15.4 communication performance does not have any big difference with and without the interference of cellular telephony and wireless networks. The lost packets indicator was high in 5 bytes transmissions over 60 m, usually with payload packets less than 10 bytes, it is not suitable for biomedical signals transmission. In a continuous transmission there is an increase in the loss of data packets when the packed size is reduced. This phenomenon is due to the capacity of management of interruptions for packet reception in the receiver device, so a lager packet sizes reduces reception interruptions and this is beneficial for the packet delivery success rate and offer less network traffic. In general, the lost packets indicator was lower with 20 bytes packets size, being slightly lower in outdoor, but only with high power and with a payload size of 10 bytes, the lost packets was lower. In average, in outdoor condition there was a 10 % of lost packets and 5% in underground. The LQI decreases with increasing distance or with obstacles in the communication, as expected. Only with high power, a payload size over 10 bytes, and with distances between 50 and over 100 m, the LQI remains near minimum value (-95 dBm), but the communications remain with a low level of losses.
M. S. Hansen, “Practical Evaluation of IEEE 802.15.4/ ZigBee Medical Sensor Networks”, M.S. thesis, Department of Electronics and Telecommunications, Norwegian University of Science and Technology, Trondheim, Norway, 2006. [2] E. Jovanov, A. Milenkovic, C. Otto, P. C de Groen. (2005, January) “A wireless body area network of intelligent motion sensors for computer assisted physical rehabilitation”, Journal of NeuroEngineering and Rehabilitation [Online], 2(6). pp. 1-3. Available: http://www.jneuroengrehab.com/content/2/1/6#B1 [3] S. Park, S Jayaraman, “Enhancing the Quality of Life Through Wearable Technology”, IEEE Engineering in Medicine and Biology Magazine, pp. 41-48, 2003. [4] B.P. Otis, J.M. Rabaey, “A 300-µW 1.9-GHz CMOS Oscillator Utilizing Micromachined Resonators”, IEEE Journal of Solid-State Circuits, pp. 1271-1274, 2003. [5] M. Ghovanloo, K. Najafi, “A BiCMOS Wireless Stimulator Chip for Micromachined Stimulating Microprobes”, Proceedings of the Second Joint EMBS/BMES Conference pp. 2113-2114, 2002. [6] Center for Wireless Integrated Microsystems (WIMS). (2010, March). [Online]. Available: http://www.wimserc.org/ [7] Zigbee Alliance, ZigBee Wireless Sensor Applications for Health,Wellness and Fitness. (2009, March). [Online]. Available: http://www.zigbee.org/Markets/ZigBeeHealthCare/Overview.aspx [8] Zigbee Alliance, ZigBee Health CareTM Profile Specification, ZigBee Profile: 0x0108, Revision 15, Version 1.0. (2010, March) [Online]. Available: http://www.zigbee.org/Markets/ZigBeeHealthCare/download.aspx [9] Digi International Inc. (2010, June). [Online]. Available: http://www.digi.com/products/wireless/zigbee-mesh/xbee-digimesh-24.jsp#specs [10] Freescale Inc. (2010, March). Simple Media Access Controller (SMAC) User’s Guide Rev. 1.5. (2008) [Online]. pp 37. Available: http://www.freescale.com/files/rf_if/doc/user_guide/SMACRM.pdf [11] Johanson Technology. (2010, March). High Frequency Ceramic Solutions. [Online]. pp 1-2. Available: http://cht.johansontechnology.com/products/rfc/ant/JTI_Antenna2450AT18A100_10-03.pdf [12] Freescale Inc. (2010, March). MC13192 2.4 GHz Low Power Transceiver for the IEEE® 802.15.4 Standard Reference Manual Rev. 1.6. [Online]. pp 106. Available: http://cache.freescale.com/files/rf_if/doc/ref_manual/MC13192RM.p df?fsrch=1
393
