Overcoming Interference is Critical to Success in a Wireless IoT World

Overcoming Interference is Critical to Success in a Wireless IoT World Ensuring reliable wireless network performance in the presence of many smart devices, and on potentially overcrowded radio bands requires a solid plan for coexistence test. As radio bands filled with wireless Internet of Things (IoT) devices utilizing different protocols 802.11, Bluetooth and others, unexplained communications failures have become commonplace, even when signal strength seems sufficiently high for good operation. This issue is especially problematic in healthcare environments where dense deployments of wireless IoT devices operating on different bands and using different radio protocols is the norm. These communication dropouts can often be attributed to several causes, in particular, issues with coexistence. Coexistence is defined as the ability of wireless equipment to operate in the presence of other equipment using dissimilar operating protocols. The only way to ensure reliable wireless network performance and in turn, succeed in the wireless IoT world especially in healthcare environments is to properly test for radio coexistence. That task is often easier said than done. WHITE PAPER Real-world example Healthcare facilities are the most challenging RF environments in the civilian world. Here s why: There are a wide range of wireless medical devices in use, as well as smart phones and wearables for personal use. Many radio protocols are incompatible and do only a marginal job of detecting other signals. Critical medical applications require complete and uninterrupted connectivity to process medical alerts and transfer large amounts of data quickly. Any disruption could negatively impact a patient s outcome Ensuring reliable wireless operation given these factors and in severely overcrowded radio conditions, makes coexistence test absolutely essential. Page 1

Why Coexistence Test? Coexistence concerns are driven by three key factors: increased use of wireless technology for critical equipment connectivity, intensive use of unlicensed or shared spectrum, and higher deployment rates of sensitive equipment like intravenous infusion pumps, pacemakers, and defibrillators. These factors directly impact how reliable medical device communications are. Coexistence problems are found in many environments. Common EMI and EMC tests only measure the proper operation of circuits against fixed standards of emissions across intended and unintended frequencies. Protocol compliance tests, such as those from the Wi-Fi Alliance, ensure that networks of similar devices using the same standard can communicate and share the channel, following the rules of that standard. However, one problem remains different standards are unable to cooperatively share channels. A Bluetooth device using frequency hopping spread spectrum (FHSS), for example, cannot detect and understand an 802.11 transmission using Orthogonal frequency-division multiplexing (OFDM) or direct sequence spread spectrum (DSSS) modulation, on the same radio frequencies. Efforts have been made to improve coexistence in these standards, but to date, no active cooperation has been defined. Bluetooth, for example, can sense which of its frequency hopping channels are experiencing a high error rate, and drop those channels from its hopping sequence. However, this indirect method of channel sharing is only marginally effective in crowded environments where many different devices operate. Coexistence Challenges Three techniques are commonly used to improve coexistence of devices and networks. Each of them face unique challenges when it pertains to healthcare environments. Co-existence testing 101 Co-existence testing measures how well a given device (or network) can operate in the presence of devices using other protocol standards on the same or nearby frequencies. Unlike compliance testing, which ensures a device/ network complies with the requirement of a given standard, coexistence testing evaluates the performance of a device/ network in a mixed-signal environment. Without appropriate coexistence testing, the reliability of the device and network in dense deployment environments, with many protocols operating simultaneously, would be highly questionable. Technique #1: Physical separation Physical separation improves network operations by reducing the signal strength of two radio networks. By placing the two networks in different locations, each network experiences a weaker signal from the other. This increases the probability that they will both be able to operate simultaneously without errors (Figure 1). One technique used to implement physical separation is known as the capture effect. In this technique, a receiver locks onto the stronger of two signals on the same frequency, while being immune to interference from weaker signals. Page 2

WiFi Hotspot Phone WiFi AP1 ZigBee 1 ZigBee 2 WiFi AP2 WiFi Client Device WiFi AP3 Figure 1: Insufficient physical separation causes interference. Challenge: This technique works well for some modulation types, but not all. As an example, an FM broadcast station may be able to use the same channel used in another town, assuming the stations are separated by 50 miles. But, this technique does not work well in healthcare environments utilizing the 2.4 GHz Industrial, Scientific and Medical (ISM) band. Although the ISM signal is much lower power, and the range is much shorter than FM broadcast, hundreds to thousands of wireless IoT devices on this band may be operating throughout the facility. For this reason, it is impractical to expect physical separation as an effective technique to reduce interference. Factor #2: Frequency separation Frequency separation is a technique used to improve the performance of mixed wireless networks. Essentially, when one network operates on a different frequency from another network, interference between the two networks is reduced, whether or not they are located close to one another. Challenge: While a valid technique, it is not always effective when it comes to the 2.4-GHz ISM band, which is occupied by overlapping Bluetooth, ZigBee, and 802.11 channels (Figure 2). Page 3

Non-overlapping channels (2.4 GHz) 802.11 a/g/n.. (WiFi) 802.15.4 (ZigBee) 3 16 79 802.15.2 (Bluetooth) Bandwidth 22 MHz 5 MHz 1 MHz Figure 2: Overlapping channels in 2.4 GHz ISM band. Bluetooth does offer one interference coping mechanism; by dropping certain channels from the hop sequence when a high error rate is experienced on those channels. This technique is a form of frequency separation. While it is a step in the right direction, it is based on the level of errors experienced by the Bluetooth device and not on intelligent agreement between the Bluetooth and Wi-Fi equipment. Consequently, it is not an optimal solution for Bluetooth networks. Factor #3: Time separation Time separation is a technique whereby transmissions are sent and received at different times to avoid collisions. This is possible because most radio networks are not transmitting 100% of the time, using only small chunk of time to transmit. As the volume of data increases; however, the radio channel is occupied more of the time, and simultaneous transmissions and collisions occur more frequently (Figure 3). Some protocols, like Wi-Fi, have collision avoidance mechanism that can detect other Wi-Fi signals and take turns using the radio channel. WiFi WiFi Bluetooth! WiFi Bluetooth ZigBee ZigBee! WiFi Time Figure 3: Different radio networks utilize different protocols, causing collisions when transmitting simultaneously. Challenge: Most radio standards are not designed to detect other network transmissions and cooperatively share channels. To make matters worse, as more data is transferred, more time is spent sending data and the corresponding acknowledgments (ACKs). This increases the chance of a device deaf to other protocols transmitting during a critical data transfer, resulting in colliding transmissions that can potentially cause errors and the need to retransmit data. Page 4

Designing the Coexistence Test Implementing one or more techniques is essential to delivering reliable device and network performance. Testing these implementations is critical to producing the highest quality device. A good coexistence test plan should include the following steps: Step 1: Characterize expected RF environment To characterize the expected RF environment, field measurements of the frequency band of interest must be performed. Because device digital transmissions are very short, a traditional swept spectrum analyzer is often ineffective for this task. The digital transmissions can come and go before the sweep even reaches the frequency in use, and thus will not be detected. Additionally, channel utilization cannot be reliably measured with a swept spectrum analyzer. For optimal field measurements: Utilize a RTSA to continually sample spectrum Perform a real-time FFT to identify the types of signals present Identify the strength of the signals present and their rates of transmission To make accurate field measurements, a Real-Time Spectrum Analyzer (RTSA) is recommended, as it allows the device designer to continually sample spectrum with a high-speed ADC. A real-time fast Fourier transform (FFT) can then be performed to convert the data into a spectral view and identify the types of signals present. To accurately model the environment for the device under test (DUT), its crucial to know what signals are present in the target environment. It may also be useful to know the strength of those signals and rates of transmission (spectrum utilization). Step 2: Choose your test signals After identifying the signals that are present, the type and number of signals needed to generate or model the coexistence test must be selected. This may mean choosing three different tiers of test signals, for example, one single Wi-Fi network passing data at the lowest tier, two Wi-Fi and one single Bluetooth signals at higher data rates, and then at the highest level, three Wi-Fi and five Bluetooth signals. Step 3: Define the functional wireless performance The next step is to identify the functional wireless performance (FWP) of the DUT. FWP is the metric used to determine success or failure of the DUT in a certain environment. It defines the important behavior required of the DUT in its radio channel. A list of the device s required functions will need to be determined, including: device startup and connection to the wireless network in the healthcare facility, successful sending of status reports, five data exchanges per minute while roaming between access points (APs), etc. The functional wireless performance requirements will depend very much on the type and application of the DUT and its defined normal operating behavior. To optimally define FWP: Understand the device type, its application and normal operating behavior Determine the required functions the device must perform, i.e. what does an intravenous infusion pump need to do to coexist? Page 5

Step 4: Choose the test s physical format There are four possible ways to configure test equipment for coexistence testing, according to IEEE/ANSI C63.27 and AAMI TIR69. Each configuration is made up of similar components: the DUT, the device that connects or pairs with the DUT, competing network devices, and a spectrum analyzer. The configuration to use depends on practical considerations, such as access to an external antenna connection on the DUT, and whether or not the device will operate in a MIMO network or whether it has directional antennas. Conducted test method This method is performed entirely with coaxial cables connecting the test equipment and the DUT. It is the most quantitatively accurate test, but the least realistic at simulating the operating environment. It is best suited for tuning the DUT hardware and firmware, and for measuring improvements resulting from the changes. To choose the right test format: Make the selection based on practical considerations Select the method that matches the accuracy you require or how accurately the method simulates the operating environment Multiple chamber test method This method uses multiple non-reflective chambers to provide a calibrated field at the location of the DUT. It allows the actual antennas for the device to be included in the test and radiated path loss to be controlled. Additionally, the shielding effectiveness of this method eliminates other potential interference sources. Radiated-anechoic chamber (RAC) test method This test method uses a single larger anechoic test chamber to ensure that the environment does not decrease the repeatability of results. This method provides additional information about the way in which the DUT fails. Assuming the pattern of the radiating antenna is known, it is possible to determine which part of the DUT is susceptible to interference. With this information, the design can be revised, mitigating the risk due to interference. Radiated Open Environment (ROE) Test Method In this open-air test method, all equipment is placed in an open area, presumably with limited active radio networks other than those used in the testing. Of all four methods, this is the least quantitative test, but it is also the most realistic. However, it is vulnerable to disruption from unexpected signals, and is therefore the least repeatable method. Page 6

Running the Coexistence Test Unlike electromagnetic compatibility (EMC) testing, coexistence test is not a pass or fail test. It focuses on the probability that the DUT can meet its FWP under the conditions of the test. For medical device tests, it provides a measure of risk that a device will or will not perform properly under the modeled conditions. Depending on the format of the test, device designers can vary the conditions of physical separation, frequency separation, data rates, and formats that are pre-defined in the FWP. As an example, a device designer might choose to vary the DUT signal strength at the receiving device from -40 to -90 dbm in steps of 5 dbm. The interfering signal can be varied over the same range of signal strength, increasing the interfering signal data rates at each step until the DUT fails to achieve the desired FWP goals. Each step test may be run for a certain period, looking for failures to meet FWP targets. The result will be a chart of Likelihood of Coexistence as defined in the test design. Regardless of the test format, the actual RF test conditions throughout the test are measured and documented using a spectrum analyzer. A Real Time Spectrum Analyzer (RTSA) that can continually sample the spectrum with a high-speed analog-to-digital converter (ADC) and then perform real-time fast Fourier transform (FFT) to convert the results to a spectral view will be needed to see the radio signals and quantify the test environment. Page 7

Conclusion With more than 20 billion wireless devices predicted to be connected by the year 2020, interference between these devices is inevitable. Coexistence test offers a way to test the ability of a device to operate in the RF environment, populated by many different wireless standards. The IEEE/ANSI C63.27 document standardizes how this type of testing is to be performed, while the AAMI TIR69 document tunes this standard for medical applications based on risk. The FDA is expected to issue recommendations for coexistence testing of wireless equipment used in healthcare soon. Regardless of the application in question healthcare, industrial monitoring and control, or environmental measurement and what radio spectrum is employed, the methods defined in these documents can be used to measure the actual performance of a wireless device in a mixed wireless environment. As the impact of interference on medical devices can be life threatening, device makers would be well advised to use one of these methods to quantify and mitigate the risk of interference. Bluetooth and the Bluetooth logos are trademarks owned by Bluetooth SIG, Inc., U.S.A. and licensed to Keysight Technologies, Inc. Learn more at: www.keysight.com For more information on Keysight Technologies products, applications or services, please contact your local Keysight office. The complete list is available at: www.keysight.com/find/contactus This information is subject to change without notice. Keysight Technologies, 2018, Published in USA, July 17, 2018, 5992-3095EN Page 8