7 Steps to Design an Effective Wireless Network in Healthcare

1. Create an Accurate Predictive Model for Wireless Networks

The predictive model is the starting point for designing a wireless network that supports healthcare system requirements for constant availability of wireless devices. A software application such as Ekahau Pro can help engineers design a wireless network. However, automated artificial intelligence modeling software alone is not sufficient for accurate radio frequency (RF) designs. Understanding the floor plan, building materials, access routes and equipment that will be connected to the wireless is crucial for building an accurate predictive model.  

Wireless engineers should start with the most up-to-date CAD drawings for the facilities where real-time patient monitoring is requested. Ideally, the wireless engineer tasked with the design will visit the site at different stages of the construction project. This allows the engineer to understand the types of materials used for the walls, ceiling and floors, as well as the materials that will be built into the facility, such as stainless steel, lead, electrical components, tube systems, plumbing, etc.

These materials introduce multipath, reflection, refraction and absorption of the RF signal into the design considerations. These components should be factored into the design of the wireless implementation for the model to be predictively accurate. The more detail incorporated into the predictive model at the beginning, the more closely the validation survey will align with the model.

RELATED: Find out how the future of smart hospital strategy brings care to the home.

2. Be Ready to Suit Up When Surveying the Healthcare Site

Surveying a construction site typically requires attending a safety class and wearing personal protective equipment (PPE) before entering the site. Alternatively, visiting an emergency department or ICU wing may require surveying outside of standard hours to avoid interfering with surgery schedules or other critical operations, in addition to wearing PPE.

During site visits, take pictures, videos and notes of the various construction stage, which will assist in adding details into the predictive model. Incorporating the knowledge of building materials, floor plans and details from photographs will result in a truly predictive model that closely aligns with all requirements when implemented. A well-designed, thorough and accurate predictive model will help reduce issues during implementation, meet business objectives in a timely manner and save the healthcare system costs in IT resources.

3. Understand the Health IT Devices Supported by the Network

The types of devices and applications that must be supported on the wireless network have an impact on the network design. It’s important to understand a device’s security capabilities, such as WPA2-PSK or WPA2-Enterprise. Typically, EAP-TLS or EAP-PEAP is used for the wireless LANs such as Vocera devices, VoIP phones, smartphones and tablets.

During the predictive modeling stages, build in the requirements using the least capable, most important device (LCMID) methodology. This is a significant challenge in healthcare environments because there are a multitude of devices constantly connecting to the wireless networks. Use LCMID methodology to focus on the critical devices and models that must connect to the network, such as patient telemetry using real-time applications, VoIP devices such as Cisco or Vocera, and mobile devices using applications such as Voalte or Mobile Heartbeat. Once these are identified, shift the focus to the vendor requirements to support these devices.

Be aware of these requirements during the modeling stages:

• The version of wireless support, such as 802.11b/g/n/ac/ax
• Channel width support: 20-40 megahertz
• Multiple-Input Multiple-Output support for the number of spatial streams 1/2/3/4
• Maximum supported transmit power
• Roaming threshold

If documentation with this information is not available, the Federal Communications Commission identifier listed on the device is an excellent resource for learning what the device can support. The FCC ID Search and Redirection tool uses the FCC ID to return this information to aid in the design of the predictive model.

LCMID vendor requirements that include newer devices capable of supporting multiple spatial streams with 802.11ac/ax are less likely to experience performance problems when connected to the WLAN or when using applications for electronic medical records, web browsers, word processors, VoIP and roaming.

Click the banner below to discover this year’s 30 health IT influencers to follow.

For example, a wireless predictive model designed for Philips Telemetry included Vocera B3000n badges for doctors and VoIP communications for nurses. In this case, the Vocera WLAN requirements and Vocera Best Practices documentation were used for this model because Vocera devices had greater WLAN coverage requirements:

• Primary Signal Strength: -65
• Secondary Signal Strength: -67
• Signal-to-Noise Ratio: 25
• Data Rate: 12
• Channel Interference: 1 at minimum Signal Strength of -85
• TX Power: Max 16 dBm (40 mW) – Min 13 dBm (20 mW)

This Philips Telemetry model was designed only for 5GHz. Unnecessary 2.4GHz radios were disabled during the design and staging of the access points. Alternatively, the wireless engineer can place disabled radios into monitor mode to optimize location tracking and security. One of the most critical steps when designing a wireless network in the healthcare environment is using 20MHz channel width.  The importance of implementing a 20MHz channel width rather than 40MHz or 80MHz channel bonding cannot be stressed enough. Utilizing a 20MHz channel width in high density deployments allows the reuse of channels to minimize co-channel interference.

A valuable predictive model design incorporates the use of 20MHz channel width and looks to achieve a high-performing 5GHz wireless network that will work reliably for the customer for the next three to five years. Soon, this requirement will shift to 6GHz. To accomplish this, there must be the ability to reuse as many channels as possible. This decreases co-channel interference in a high-density deployment. Vocera provides detailed documentation of the requirements for a successful, high-performing wireless network.

DISCOVER: How Cisco’s OpenRoaming rollout enables quick, secure Wi-Fi access at Adventist Health.

4. Prepare to Deploy a Predictive Model for Wireless Networks

Prior to starting a predictive model, set three or four Vocera badges connected to a WLAN next to each other. Next, open the badge settings to view the Received Signal Strength Indicator (RSSI) reported by each badge. Then, connect an Ekauhau Sidekick device to the Ekahau Pro spectrum analyzer software to view the measured RSSI for that area. Finally, determine the difference between the highest RSSI and the lowest RSSI measured by the badges and the Sidekick device. This difference is the offset for the wireless design.

For example, say the badges have the following RSSIs:

• Badge 1: -70 dB
• Badge 2: -72 dB
• Badge 3: -75 dB
• Badge 4: -68 dB
• Sidekick device viewing as measured: -65 dB

The difference of the highest measured RSSI and the lowest measured RSSI results in an offset value of -10 dB. Therefore, with a built-in offset, the primary coverage is -55 dB and the secondary coverage is -57 dB. Ekahau Pro modeling software is now able to compute this with reliable accuracy using the “View as Mobile Device” feature.

5. Create a Wireless Network Installation Guide

Once the predictive model design is complete, a list of the hardware required to implement the design should be developed. Generating a comprehensive guide listing the type of access points, external antennas and other components used assists with creating the bill of materials and an instruction guide for the installation vendor. All access points, external antennas and other necessary hardware should be included in the bill of materials and the installation guide. The guide documents the following information at minimum:

• Access point model
• Antenna type
• Hardware
• Placement
• Height
• Angle

The guide provides clear instructions for the installation and placement of each access point and antenna.

It is also beneficial to document the access point name, physical location, MAC address and radio MAC address (for Vocera purposes) to enter the Vocera Administration server or Cisco Prime Infrastructure/DNA Center in the proper location on the maps.

EXPLORE: 4 key advantages of SD-WAN technology for healthcare.

6. Prepare for Wireless Network Installation

Prior to installation, prepare the access points for deployment by connecting Power over Ethernet (PoE) to the access points and registering them to the proper wireless controller. On the wireless controller, create new (if needed) RF profiles for both the 2.4GHz and 5GHz radios along with an AP group. Add the access points to the AP group at this time. Radio resource management (RRM) settings allow wireless engineers to be very specific with radio configurations. It is best not to leave the default settings in place. Customizing the settings allows the engineer to optimize the performance and reliability of the wireless network.

In the aforementioned Philips Telemetry model, the channel plan used for the wireless design turned off unnecessary 2.4GHz radios or put them into monitor mode. Changes to the design were made to accommodate the floor, building or location, and to support the type of client devices in use. 

7. Install the Wireless Network Hardware

It is recommended that the engineer conduct onsite visits during the deployment process to verify the proper installation and the associated locations of the access points. Access points should not be hanging from a ceiling tile in a way that points the RF in an incorrect or inefficient direction. Care should be taken to correctly mount access points in an aesthetically pleasing manner, especially in public spaces and new buildings.

Once the access points have been installed and the locations verified, they can be powered on. After they have been powered on, allow the access points to run for at least 24 hours. This allows RRM to make changes to the power and channel settings and other configurations.

Finally, the wireless network is deployed, powered on and operating. A validation survey should now be performed to evaluate how the operational network aligns with the designed predictive model. The results of the validation survey provide actionable feedback to improve the wireless network, such as adding additional access points, moving access points or simply increasing or decreasing the transmit power on specific access points. It’s best practice to have at least a 15-foot service loop so that access points can be moved easily. RRM algorithms are not completely accurate and, at times, it’s necessary to manually set access point channels and power.

These adjustments optimize the wireless network within the physical space. In healthcare, and any environment where reliability and strong signal are paramount, a second validation survey should be conducted. This survey evaluates the effectiveness of the changes made to the wireless access points as a result of the first survey. This second survey validates the implementation outcome against the predictive design and that the network meets all the requirements for the types of devices it will be supporting.

Strong, reliable wireless networks are part of the critical infrastructure of a hospital. Take the time to encompass all facets of the scenario in the predictive model. Listen to doctors, nurses and other customers to understand what they need and the types of devices they use. Be present and actively involved during all phases of a build to adjust the model for building materials. Software applications and AI are good tools, but they are just that: tools. Excellent design requires an engaged human engineer to plan, adjust and compensate for the complex components that are included in patient telemetry. Take the time to collaborate with vendors and embrace proper design protocols. In the healthcare environment, lives depend on a well-planned network.