Tests to assess the risks associated with coexistence of wireless technologies are necessary for safe and effective RF wireless medical devices.
At any given time, a typical home or hospital uses a number of wireless systems (e.g., IEEE 802.11a/b/g/n, or WiFi; Bluetooth; ZigBee; cordless phones) operating on the same industrial, scientific, and medical (ISM) band.1,2 Given the increasing use of wireless, RF wireless medical devices and other wireless systems operating nearby can interfere with each other. If a collision between their respective transmissions occurs, data packets transmitted by medical devices could be delayed or blocked, potentially interfering with timely transmissions of critical data. Techniques such as retransmission and forward error correction might no longer be sufficient to overcome interference and spectrum congestion. Hence, methods to design and test wirelessly enabled medical devices for risks associated with coexistence of wireless technologies are essential for innovative, safe, and effective RF wireless medical devices.
Although there is some overlap between electromagnetic compatibility (EMC) and wireless coexistence, differences exist. Wireless coexistence is the ability of one wireless system to perform a task in an environment where other systems that may or may not be using the same set of rules can also perform their tasks.3 EMC is the ability of a device to function properly in its intended electromagnetic environment without introducing excessive electromagnetic energy that could interfere with other devices. Manufacturers of electrically powered medical devices routinely test their equipment to applicable national and international consensus safety standards. EMC test results are often used to support safety claims to regulatory agencies such as FDA. Less well-known are the issues and concerns associated with wirelessly enabled medical devices, although this is changing thanks to FDA’s guidance document on wireless medical devices.4
To date, no consensus standards adequately address the risks associated with wireless coexistence for medical devices and systems. Current methods of evaluating wireless coexistence use ad hoc test methods that vary widely among device manufacturers and test facilities. Moreover, current medical device EMC standards have no requirements or test procedures to assess the performance of systems containing RF receivers in the presence of in-band transmitters. This article examines the limitations of present medical device EMC standards for coexistence evaluation, identifies factors to be examined when testing for coexistence, and discusses the status of plans to develop a wireless coexistence test method.
International Electrotechnical Commission (IEC) 60601-1-2— a collateral to IEC 60601-1, the general safety standard for medical electrical equipment—is the primary standard used for EMC testing of nonimplanted and nonin-vitro diagnostic electrical medical devices.5,6 Even if the wireless technology is considered part of the essential performance and function of the device, meaning its absence would result in unacceptable risk, the testing and requirements in these standards and the related documents do not fully address the characteristics and performance associated with the wireless technology.
One major limitation of the IEC 60601-1-2 standard is that RF receivers are exempt from immunity testing in the exclusion band (passband). This exemption is granted because a test signal in the passband of a traditional RF receiver would be expected to cause interference. The exclusion band is defined in 3.10 of IEC 60601-1-2 as the:
…frequency band for intentional receivers of RF electromagnetic energy that extends from -5 % to +5 % of the frequency, or frequency band, of reception for frequencies of reception greater than or equal to 80 MHz and from -10 % to +10 % of the frequency, or frequency band, of reception for frequencies of reception less than 80 MHz”
For a wireless medical device system operating at 2.45 GHz, the exclusion band covers the entire 2.4 GHz ISM band. In other words, a medical device with RF wireless technology is not required by IEC 60601-1-2 to be able to maintain wireless communication when subjected to RF signals in its passband. Therefore, as published, this standard does not provide means for assessing the performance of the medical device wireless communication system. A new draft of Edition 4 of IEC 60601-1-2 would require that medical equipment remain safe (i.e., provide basic safety and essential performance) when exposed to an RF immunity test signal in the passband but still would not assess nonsafety performance under these conditions.7
Wireless coexistence can also be defined as the ability of multiple wireless systems to share the same or adjacent frequency spectrum without undue interaction or interference affecting performance and transmission or reception of signals and data. IEEE 802.15.2 discusses computer modeling and design issues for improving wireless coexistence between IEEE 802.15 and IEEE 802.11 communication systems.8 The document does not, however, specify a method or pass-fail criteria for testing coexistence, leaving a gap in the ability to assess the characteristics and performance of wireless systems.
Due to concern for patient safety and medical device effectiveness, FDA drafted a guidance document to assist in the design and testing of RF wireless technology in medical devices.4 The document calls attention to general risks and concerns for devices operating in the crowded RF spectrum—particularly in the ISM frequency bands. Other risks, including those that might be affected by quality of service, data integrity, wireless security, and EMC of wireless technologies and medical device functions, are also outlined. In addition, the document provides information to assist in preparing regulatory submissions. FDA remains active and has an interest in the development of standards and evaluation methods for determining and validating the performance of RF wireless technology in medical devices in general and wireless coexistence in particular.9
The following are major factors influencing the coexistence of wireless medical devices operating in the presence of other heterogeneous wireless networks, including radio channel characteristics, antenna and signal polarization, frequency bands, separation distances, and cochannel and adjacent-channel interference. These factors should be considered when developing a standardized test protocol for RF wireless coexistence.
Radio Channel Characteristics. The environment in which an RF wireless medical device is evaluated for coexistence is critical and must be well-characterized to mimic the expected environment and typical deployment. Medical device deployments include those that can be connected to or implanted in stationary or mobile patients. Device deployment, configuration, and environment affect the wave propagation through the wireless channel. Two such wave propagation scenarios, line-of-sight (LOS) and nonline-of-sight (NLOS), must be examined. LOS testing can be performed in an anechoic chamber, ensuring reproducibility by isolating the medical device and its wireless technology from spurious interference. Conversely, NLOS testing is performed outside an anechoic chamber to account for the effects of reflection-caused multipath on the received RF signal.
Performance in an NLOS deployment depends upon radio channel path loss among the transmitter and receiver, frequency band, and transmission power. In an NLOS test setup, as the signal-to-interference ratio of the wireless medical device decreases, the likelihood of interference from nearby wireless systems increases.10,11 The delay in the traffic flows is also exponentially related to the signal-to-noise ratio of the wireless medical device network under test, which would cause an increase in latency for the wireless medical device data.12,13
This type of testing for the NLOS configuration outside of an anechoic chamber raises the possibility of collateral phenomena, including reflections from nearby structures causing multipath. To account for this, multiple test arrangements of the transmitter and receiver terminals should be considered. For each test setup, the wireless channel can be characterized by finding its path loss attenuation and power delay spread, improving test result reproducibility from one testing environment to another.
The wireless medical device transceivers should be evaluated separately (see Figure 1)—each separately exposed to one or multiple interfering wireless networks. Various interference phenomena arise depending on whether the interfering wireless network is in the proximity of the wireless medical device transmitter or receiver. When a medical device receiver is surrounded by an interfering network or networks, packet collisions increase at the receiver (i.e., the hidden terminal effect). In contrast, when a medical device transmitter is surrounded by one or more interfering networks, channel utilization decreases (i.e., the exposed terminal effect). Decreased channel utilization is the result of busy channel sensing followed by increased backoff (contention) windows.
Polarization. Polarization and cross-polarization of both the wireless medical device and interfering terminals should be considered when testing for coexistence. For example, if the RF fields from the wireless medical device are intended to be horizontally polarized when operated, the interfering network should be deployed in both horizontal and vertical polarizations because an antenna is never 100% polarized in a single mode. Additionally, radio channel multipath could cause cross polarization in an NLOS test environment.
|Figure 1. LOS Test Setup for Wireless Medical Device Telemetry System|
Cochannel and Adjacent-Channel Interference. Cochannel and adjacent-channel interference must be considered to effectively characterize the probability of packet collision.14,15 The effect of adjacent-channel interference on the bit-error rate has been shown to strongly depend on the frequency offset between the channel under test and the interfering carriers.16,17 The correlation between spatial distance and channel spacing to control interference between concurrent transmissions in a wireless sensor multichannel network has also been investigated.18 The impact of cochannel interference was tested experimentally, and performance was analyzed for IEEE 802.11g networks.19 The impact of adjacent channel interference in 802.15.4 (ZigBee) networks has also been studied.14 Coexistence testing was performed with a single interfering terminal and multiple interfering terminals in different combinations, including two interferers in the same adjacent channel, two interferers in two different adjacent channels, and three interferers in the same adjacent channel. It has been suggested that the key to determining coexistence in heterogeneous networks is the study of the effects of multiple interferers.20
While a number of studies were limited to a specific wireless technology, the aforementioned practices are recommended across various wireless technology platforms. Hence, to evaluate a medical device for RF wireless coexistence, testing should consider a pair (server,client) of interfering terminals configured to operate on the cochannel and subsequently on the two immediate adjacent channels of the wireless medical device under test. Does one pair of interfering terminals operating on the cochannel frequency of the medical device present the worst-case scenario? The answer depends on the duty cycle (i.e., packet generation and transmission rate) of the interfering terminals. If the duty cycle is high, the cochannel will be proportionally highly utilized. If multiple pairs of interfering terminals are configured to operate on the cochannel, the exposed terminal problem arises, resulting in cochannel underutilization.
Most wireless technologies (e.g, WiFi and ZigBee) employ carrier sensing multiple access (i.e., CSMA/CA protocol) to avoid packet collision. This regulates channel accessibility by permitting one terminal to transmit at any given time. Testing with multiple interfering terminals—where one pair of terminals is set on the cochannel frequency and two others are set on two different adjacent channels—is highly advantageous, as it evaluates performance of medical devices designed to automatically select an alternate channel when interference exists on its current channel. Testing will evaluate the accuracy of the new selection and the time required for the medical device to switch to the new channel.
|Transmitter Power||Initial Distance m||
Minimum Distance m
|600 mW–2 W||2||0.5|
|2 W–8 W||3||1|
Distance. The distance between the interfering wireless terminal or network and the RF wireless medical device transmitter or receiver requires consideration because the likelihood of interference with the medical device increases as the signal-to-interference ratio decreases.10,11 Adjusting the distance between the medical device transmitter or receiver and the interfering wireless terminal or network is one way to control the signal-to-interference ratio. The initial and minimum test separation distances recommended by ANSI C63.18 should be used, as shown in Table 1.21 Initial and minimum distances are determined based on the transmitting power of the interfering terminals. The initial test distance is calculated to expose the device under test to approximately 3 V/m. The minimum test distances were calculated so that the device under test would be exposed to approximately 20 V/m. In the C63.18 test method, if there is no interference at the initial distance, the distance between the wireless medical device receiver and interfering networks is systematically reduced until interference occurs or the minimum test distance is reached. Auto-power-leveling algorithms should be disabled in interfering terminals. Allowed maximum power should be configured for each interfering wireless device used. The C63.18 test procedure recommends configuring transmitters to maximum output power. If interference is present at the initial distance, then the distance between the medical device terminal and the interfering network terminals is increased until interference ceases. Results are most reproducible when the output power of the interfering terminals does not change during the test. The results obtained at various separation distances are used to characterize the wireless medical device and inform policies and procedures for EMC management in healthcare facilities.
The transmission parameters of the wireless medical device (e.g., packet size, polling window, clear channel assessment threshold, and duty cycle) can alter the outcome of coexistence testing. Studies have shown that as the packet size increases, the probability of packet loss increases.10,11,22 Studies have also shown that as the polling window increases, the probability of packet loss decreases.11,22 Additionally, results indicate that when the interference level is below a certain level specified by the device sensitivity, the channel is sensed as idle, or clear, and the interference does not affect communication.10,11,14,22 Duty cycle, or channel utilization, is mainly dependent on the amount of traffic generated and transmitted by the interfering wireless networks. Studies show that as the interfering device or network increases its duty cycle, the victim network packet loss ratio increases, causing either temporary or permanent interference.10,11,22 Therefore, two transmission parameter settings should be used during coexistence testing: typical or manufacturer-suggested default settings and worst-case-scenario settings (e.g., as suggested by previous work published in the literature).
Assimilating the aforementioned coexistence factors into a test protocol is dependent upon the medical device under test. A case study is offered to serve as an example. Of note is that a test setup is unique to a specific wireless medical device or system.
The wireless medical system under test was a telemetry system implementing a ZigBee chipset to establish bi-directional communication between two parts of the system, hereafter designated as device #1 and device #2. ZigBee is a low-power communication protocol with a bit rate of 250 kilobits per second in the 2.4 GHz band. When using ZigBee, selecting channel 13 or 21 for coexistence testing is suggested, the justification for which is explained below. Channel 21, i.e., 2.455 GHz, was used by the medical wireless system for this case study.
The RF wireless medical device system was first baselined under static conditions without interference. The test layout is shown in Figure 1. Medical devices were separated by a distance of 5 m and elevated on a wood table to a height of 1 m. The separation distance and height were chosen based on the typical deployment of the wireless medical telemetry system.
The wireless medical system transmission parameters were evaluated using two configurations: typical settings (i.e., default manufacturer settings) and worst-case-scenario settings, as explained previously. The medical device system communicated a predetermined number of packets to establish a statistical baseline of the wireless link for each transmission parameter setting. To characterize the wireless medical telemetry system for this case study, the measured parameters consisted of packet loss ratio, delay, and throughput. These parameters are used to assess the quality of service provided by the wireless medical device. The desired quality of service is dependent upon the critical functions specified by the medical device manufacturer.
After establishing a baseline, the wireless medical device system was evaluated for coexistence with other wireless technologies. An IEEE 802.11g network, e.g. Wi-Fi, was deployed around medical device #2. The 802.11g network consisted of up to three pairs, each operating on a different channel. It should be noted that auto-power-leveling algorithms are disabled in the 802.11g network, and the maximum power allowed by the wireless network standard is used. The initial separation distance between the medical device and the interfering network was determined based upon the transmission power of the interfering network, which is less than 600 mW. The ANSI C63.18 table (Table 1) suggested that the initial distance should be 1 m.
For the first set of coexistence tests, only a single 802.11g network was introduced to the wireless medical device network. Communication was first established between the wireless medical devices and then in the 802.11g network. In an attempt to cause interference with the wireless medical device, the testing parameters were: channels 1-11, data rate of 1-26 Mbit/s, horizontal and vertical polarization, variable separation distance between the medical device and interfering network and a single 802.11g interfering network. After an adequate number of packet transmission attempts between the medical devices, the interfering 802.11g network was turned off. If interference occurred during testing, the amount of time was determined that the wireless medical device telemetry system needed to restore communication after the cause of interference was removed. After all variables were tested, the positions of medical device #1 and medical device #2 were exchanged, and testing was repeated so that each medical device was tested separately with the interfering network.
The RF wireless medical device system was then exposed to three 802.11g wireless networks simultaneously. Channels 9, 10, and 11 were chosen. The channel selection allows co-channel interference (10) and two different adjacent channel interferences (9, 11) to be tested for the wireless medical device telemetry system. The data rate, separation distance, and polarization of the three 802.11g networks were variables that were evaluated. Again, medical device #1 and medical device #2 were separately exposed to the interfering network.
The NLOS test setup was then performed. The experimental methodology for LOS and NLOS testing was the same, i.e., the RF wireless medical telemetry system was first baselined without an interfering network and then interfering networks were introduced into the environment.
A LOS and an NLOS test setup were then repeated with different interfering networks in the same frequency band as the wireless medical telemetry system. Examples include ZigBee, IEEE 802.11b, IEEE 802.11g, IEEE 802.11n, Bluetooth, and cordless phones, which transmit at 2.4 GHz.
The wireless medical network did not experience a loss of network connectivity during the LOS testing with one or three 802.11g wireless networks operating simultaneously in the same unlicensed spectrum band at a separation distance of 0.25 m between the interfering networks and the wireless medical device under test. All of the wireless functions operated as intended.
The wireless medical network experienced a loss of network connectivity during the NLOS testing when one or three 802.11g wireless networks were operating simultaneously in the same unlicensed spectrum band with a separation of 3 m or less between the interfering networks and the wireless medical device under test. When the separation distance was greater than 3 m, network connectivity was restored to the wireless medical network. When the throughput (duty cycle) of the 802.11g network was decreased, network connectivity was restored to the wireless medical network at a separation distance of 1 m. It was also observed that shorter wireless medical device packets had a higher probability of successful transmission while longer data packets had a lower probability of successful transmission.
The RF wireless parameters discussed and the recommendations made in this article are intended to serve as a starting point for medical equipment manufacturers and healthcare organizations in assessing the coexistence of RF wireless communication in modern dynamic electromagnetic environments. Coexistence testing is different from EMC RF immunity testing, under which the RF immunity of a medical device is characterized over a range of frequencies.
Pass-fail criteria for coexistence testing must be identified and need to be quantified in terms of the critical functions of the device under test as determined by risk assessment. The following parameters can be considered:
In other words, coexistence testing should demonstrate that the effects of the wireless medical device on nearby RF wireless equipment and networks is minimal and effects of nearby RF wireless equipment and networks on the functions of the wireless medical device would not result in unacceptable risk to the patient or user.
Several groups interested in RF wireless coexistence testing of wireless medical devices have begun work in this area. A working group of ANSI-accredited committee C63 held a preliminary teleconference in January 2011, and FDA sponsored a teleconference in February 2011. The University of Oklahoma Wireless and Electromagnetic Compliance and Design Center, in Tulsa, has an evolving test protocol based on the parameters listed. The University of Oklahoma is currently experimentally testing medical devices for coexistence. Plans for future work include proposing that ANSI-accredited committee C63 establish an official working group to develop consensus coexistence standards or recommended practices.
1. N. LaSorte, Y. Burette, H. Refai, “Experimental Characterization of Electromagnetic Propagation of a Hospital from 55-1950MHz”, in Proceedings of IEEE Asia-Pacific Symposium on Electromagnetic Compatibility, April, 2010, pp. 826 – 829.
2. N. LaSorte, W. Barnes, H. Refai, “Characterization of the Electromagnetic Environment in a Hospital and Propagation Study”, in Proceedings of IEEE International Symposium on Electromagnetic Compatibility, August 2009, pp. 135-140.
3. Institute of Electrical and Electronics Engineers. IEEE Recommended Practice for Information technology—Telecommunications and information exchange between systems — Local and metropolitan area networks — Specific requirements, Part 15.2: Coexistence of wireless personal area networks with other wireless devices operating in unlicensed frequency bands. IEEE 802.15.2:2003. New York, NY: IEEE, 2003.
5. International Electrotechnical Commission. Medical electrical equipment—Part 1: General requirements for basic safety and essential performance. IEC 60601-1:2005. Geneva (Switzerland): IEC, 2005.
6. International Electrotechnical Commission. Medical electrical equipment—Part 1: General requirements for basic safety and essential performance. 2. Collateral standard: Electromagnetic compatibility—Requirements and tests. IEC 60601-1-2:2007. Geneva (Switzerland): IEC, 2007.
7. International Electrotechnical Commission. Medical electrical equipment—Part 1: General requirements for basic safety and essential performance. 2. Collateral standard: Electromagnetic disturbances—Requirements and tests. IEC 62A/746/CD, draft Edition 4 of IEC 60601-1-2. Geneva (Switzerland): IEC, 2011.
8. Institute of Electrical and Electronics Engineers. IEEE Recommended Practice for Information technology — Telecommunications and information exchange between systems — Local and metropolitan area networks — Specific requirements, Part 15.2: Coexistence of wireless personal area networks with other wireless devices operating in unlicensed frequency bands. IEEE 802.15.2:2003. New York, NY: IEEE, 2003.
9. D. Witters, S. Seidman, H. Bassen, “EMC and Wireless Healthcare,” IEEE 2010 APEMC, 2010.
10. L. Angrisani, and M. Vadursi, “Cross-Layer Measurements for a Comprehensive Characterization of Wireless Networks in the Presence of Interference,” IEEE Trans. Instrum. Meas., vol. 56, no. 4, pp. 1148-1156, Aug. 2007.
11. L. Angrisani, M. Bertocco, D. Fortin, and A. Sona, “Experimental study of coexistence issues between IEEE 802.11b and IEEE 802.15.4 wireless networks,” IEEE Trans. Instrum. Meas., vol. 57, pp. 1514–1523, 2008.
12. Tianlin Wang, and Hazem H. Refai, “The Development of an Empirical Delay Model for IEEE 802.11b/g based on SNR Measurements”, Proceedings of the 1st IEEE International Symposium on Wireless Quality-of-Service, Maui, Hawaii, June 13-16, 2005.
13. Tianlin Wang, and Hazem H. Refai,” Empirical Network Performance Analysis on IEEE 802.11a/b/g with Different Protocols and Signal to Noise Ratio Values”, Proceedings of the IEEE International Conference on Wireless and Optical Communications Networks (WOCN 2005), Dubai, UAE, March 6-8, 2005.
14. L. Lo Bello, E. Toscano. “Coexistence Issue of Multiple Co-Located IEEE 802.15.4/ZigBee Networks Running on Adjacent Radio Channels in Industrial Environments,” IEEE Trans. Industrial Infomatics, vol. 5, pp.157-167, 2009.
15. I. Howitt, “WLAN and WPAN coexistence in UL band,” Transactions on Vehicular Technology, vol. 50, pp. 1114–1124, 2001.
16. J. C. Haartsen, “The bluetooth radio system,” IEEE Personal Communications, vol. 7, pp. 28–36, 2000.
17. A. Kamerman, “Coexistence between Bluetooth and IEEE 802.11 CCK solutions to avoid mutual interference,” IEEE 802.11-00/162, 2000.
18. O. Incel, S.O. Dulman, P.G. Jansen, “Multi-channel support for dense wireless sensor networking,” in Proceedings 31st IEEE Conference on Local Computer Networks, November 2006, pp. 694-701.
19. G. Cena, I. Cibrario Bertolotii, A. Valenzano, C. Zunino, “Reasoning About Communication Latencies in Real WLANs,” in Proceedings 12th IEEE Conference on Emerging Technology Factory Automation, September 2007, pp. 187-194.
20. P. Pinto, A. Giorgetti, M. Chiani, and M. Win, “A stochastic geometry approach to coexistence in heterogeneous wireless networks,” IEEE Journal on Sel. Areas in Communications, vol. 7, pp 1268-1282, 2009.
21. Institute of Electrical and Electronics Engineers, Inc. American national standard recommended practice for on-site ad hoc test method for estimating radiated electromagnetic immunity of medical devices to specific radio-frequency transmitters. (standard C63.18). Piscataway, NJ: IEEE, 1997.
22. M. Bertocco, A. Sona, F. Tramarin, “Design of Experiments for the Assessment of Coexistence Between Wireless Networks,” in Proceedings of the Instrumentation and Measurement Technology Conference, May 3-6, 2010, pp. 928-932.
Nickolas LaSorte is pursuing a PhD in electrical engineering at the University of Oklahoma. Hazem H. Refai is an associate professor in the university’s school of electrical and computer engineering and founding director of the WECAD Center. Seth Seidman is a research electrical engineer at FDA. Donald Witters is chairman of the CDRH EMC and Wireless Group. Jeffrey L. Silberberg is senior electronics engineer for CDRH, Office of Science and Engineering Laboratories, Division of Electrical and Software Engineering.