The Healthcare and Enterprise Wireless Experience

802.11n 802.11a/g Multipath the Site Survey and Beamforming

From a healthcare perspective (via any vertical market) the hospital is the worst multi-path environment available. This multipath environment is caused by all the reflective elements present such as steel carts, beds, fluids, and tile/reflective surfaces, plus everybody moving around 24/7/365.  Add to this all the movement and mobility that is constantly going on as described.  Not a good environment for a RF signal to play in..period.  Example your smart phone in the depths of building. This is bad for the performance of 802.11a/b/g, but actually is an advantage for 802.11n.  Understand that MIMO with 802.11n, takes advantage of multi-path.  So then how has the site survey changed and what is the true throughput of 802.11n?  The traditional site survey (for WLAN) plans for the performance of negating multi-path with 802.11a/b/g, but this will degrade the potential performance of 802.11n, thus decreasing available network throughput. This requires a careful thought process "and" site survey on how to balance both what is somewhat considered legacy WIFI (802.11a/b/g)with 802.11n. (Let me say that 802.11a/b/g is not legacy, but the IT mindset is focused on speed...more better) Just remember it is not always about speed, but for the application to perform in the correct defined business model.  In this case a lot of medical device applications with 802.11a/b/g work extremely well.  Add to this, not many if limited OEM embedded 802.11n modules are offered today.  This will mature, just like 802.11b to 802.11a/b/g, and 802.11n, once power mangement issues are resolved regarding MIMO. (It takes a lot of power to transmit and receive). Sorry it is physics.

Beamforming is a great "marketing term", sounds like something dreamed on a napkin over a few bottles of beer or glasses of wine, but in fact it is pretty simple and has been around the the wireless world for a long time.  Beamforming is pretty straight forward. It is method of directing RF energy in order to improve the signal to noise ratio (SNR) at the receiver, thus improving the reliability of the signal.  When you have more antenna elements (dynamic beamforming) to select a signal in an x/y/z axis you then can use software and algorithms to clean up the signal relative to the noise floor and then send back this signal in phase using the same antenna elements.  So, what does this all mean? OK, Cisco uses 'beamforming", and also Ruckus Wireless. I go back to my old days of "digital radio receivers".  If you use 14 to 3 antenna elements you tend to receive a better S/N of which is based upon all the same output of the transmitter and received receiver. It also probably overcomes a lot of the noise floor issues, since you have more antenna elements to receive and thus transmit.  Thus the algorithms can clean up the signal... in fact have a "cleaner signal", to transmit back.  Being able to look at the RF signal in a 3D axis, i.e. x/y/z, will in addition provide the value add of discerning the holistic environment from an actual multi-path and real life(demonstrated)  RF performance venue provides major league value to ensure the ultimate network throughput.  802.11n is (each end point entity), wlll need to evaluate whether DAS can fit within a financial or technical within this broadband model..Enough said.  Just look to those business models that tried a go at this!  It is my opinion that RF performance not only needs to be looked at from the network valuation; but how the RF design will utimately impact the back end network "performance" model.  While it is known that players like www.ruckuswireless.com have little footprint (at least in healthcare), their technology and performance just makes it seems some of a technical and financial sense against the yeah what network dudes consider. www.cisco.com. (I am not biased but just reading the performance RF models, of the leading network wireless performance modeling...leave it at that!)  However RF is very complex and the more and more push to the RF  envelope making it 100% 24/7/365  simply means better and confirmed absolute performance. Cisco Systems (while I got to digress they did not get where they are with not a lot of brilliant people), it just holds that a lot of smaller players with innovative technologies are nipping at their ankles). Heck, just think of Apple versus Microsoft! They the global leader in networking  www.cisco.com tend to inovate by  past experience via acquisition when competition threatens; so stayed tuned to the next venue of the next stage of wireless..it will be very interesing.  They (Cisco Systems) will and have always kept their eye on the ball of technology innovation. Who is the next target for acquistion?

 

 

 

 

WLAN Interference Rejection for Complex Healthcare Environments

Healthcare is the worst multi-path environment for RF. This is just not me saying this, it is because I had to design antenna systems for VHF telemetry and make stuff actually work!  The problem is all the fluids, metal, and moving objects, but now comes the harder issue.  The majority of wireless LANS were designed for primarily data and voice.  And you have to look at how they are being used for healthcare for the EMR (laptop on wheels, note I am not using the outdated term!), and voice over IP.  However, the laptop is pretty much outside in the hallway, “occasionally” taken into the patient room, and the clinician is using the Vocera,  ASCOM, or Cisco Systems “wireless” voice over IP phone…in the hallway.  As more and more medical devices, pulse oximeters, infusion pumps, patient monitoring are incorporating WiFi, (802.11a/b/g), they will tend to be with the patient. This by virtue of leads, IV tubing that are connected to the patient.  At times they will be walking the hallway, but the majority of times, they will be in bed in the patient room.  Now here comes the challenge.  The AP(s) are the hallway and were site surveyed for data and voice and maybe the medical device application.  Did the site survey taken into account how the signal has to penetrate the firewall (between hallway and patient room), and the glass, steel, and ceramic tile of the bathroom that is often between the hallway and the patient bed where the WiFi enabled medical device is?  Did the site survey take into account these interferers?  

Dealing properly with interference requires something be done to the RF energy coming out of the AP. You need to have complete control over it. The paths that Wi-Fi transmissions take need to be clean. And there needs to be enough paths available to you so if the one you picked is not performing, you can pick another one....REALLY fast. (Hence dealing potential drop-out) And all this needs to happen automatically.  So how does the system know it's picking the right path to the station? Answer: the client can tell you. Clients talk back, telling APs the throughput, packet error rate, RSSI and other things.

So here are some of the issues:

• 2.4GHz only has 3 non overlapping channels so there's only so much changing that can be done.
• Running 40MHz wide channels in the 5GHz band gives you only 11 non-overlapping channels, same problem.
• Changing channels changes everything (over and over)
• RF interference avoidance from the perspective of the APs is just that, the perspective of the AP. Maybe the client sees something different and usually always does. Now what are you going to do?
• Cisco's radio resource management still decides what channel to use with CleanAir.  Many Cisco customers and partners advocate turning RRM off, just Google "RRM problems" Ultimately to fix the RF interference problem it comes down to using a "purpose-built" (Cute meaningless marketing phrase) product - specifically designed with forethought to deal with, and to automatically mitigate, interference - all without network managers having to do anything!  Again, I go back to hospital construction.  Not only are their firewalls and glass and metal between the patient rooms and the hallways, but many ICU/CCU areas also have isolation rooms, or a lead lined room for fluoroscopy (X-Rays).  Did the site survey take this into effect?  Finally what about how the signal get’s through old construction using chicken wire lath, or new construction with poured concrete metal pans between floors.  All this potentially now dictates the use of “smart antennas”.

That's the real beauty of smart antennas and dynamic beamforming.

Smarter approaches to RF optimization in the form of dynamic beamforming solve some of the CORE Wi-Fi challenges facing the industry:

• Virtually eliminating co-channel interference
• Enabling more predicable performance by "steering" signals over the highest performing paths.
• Doubling the effective footprint (coverage area) of an AP (read: fewer APs give better performance).  (NOTE: Adding more AP(s), think of for location tracking (RFLS), only “increases” co-channel interference.  So think of a dense AP environment, with challenging physical construction and a difficult application environment.

The goal of beam forming is to improve the SNR, (signal to noise ratio) provide higher data rates, improve reliable connectivity, and decrease extraneous signals. It should ideally use dynamic antenna array with switching logic.

So how does this dynamic beamforming work?  Smart antennas steer Wi-Fi signals over the best paths to the client, (in this case the medical device), while the client moves the signal path is dynamic and it switchs thus nullifying the sources of interference. The combination of antenna diversity, explicit signal path control and multipath, all to steer signals around all this stuff in the hospital environment.

And what are the benefits?  Forms and steers Wi-Fi clients over the best performing signal path on a per packet basis, while using multipath to it’s advantage.  The client MAC ACKS are used to ensure delivery at the best data rate with the lowest packet loss. This all translates into automatic adaptation to real time change in the healthcare environment and provide increased signal coverage with 2 to 4 times less APs, while finally simplifying RF planned. Lower CAPEX, OPEX, and TCO.  Oh, for those that think WLAN should be on a DAS, (I never was an advocate, even though it sounded like cute marketing KoolAid), how can you take advantage of intelligent beamforming (antenna array), on a DAS whereby diversity was never used for cellular and/PCS.  OK, SISO eventually LTE.

For those that want to delve deeper into this a “black paper” is attached.

 

Download Wireless_network_perf_bp