Wireless Access

Just got off the phone with a fellow engineer in the UK. It was a conversation around  what survey model are engineers leading with today. I thought this was a great topic for a forum post to get a sense what everyone is using ? 

Active Survey - Whereby mapping wifi coverage with an ap on a stick approach, which is very labor intensive, but much more accurate. 

Blue Print Survey (predictive) - Whereby you take a floor plan and use software to map the planed area of coverage and then follow up with a passive survey when deployed.

Personally I use a hybrid of both. All depends on the need. If its voice, I almost always do active, I just don't want any surprises. I do predictive if the budget can't justify the cost of active or the area of coverage is needed for data only. 

What do you lead with and why?

---

@gstefanick wrote:

Just got off the phone with a fellow engineer in the UK. It was a conversation around  what survey model are engineers leading with today. I thought this was a great topic for a forum post to get a sense what everyone is using ? 

 

Active Survey - Whereby mapping wifi coverage with an ap on a stick approach, which is very labor intensive, but much more accurate. 

 

Blue Print Survey (predictive) - Whereby you take a floor plan and use software to map the planed area of coverage and then follow up with a passive survey when deployed.

 

Personally I use a hybrid of both. All depends on the need. If its voice, I almost always do active, I just don't want any surprises. I do predictive if the budget can't justify the cost of active or the area of coverage is needed for data only. 

 

What do you lead with and why?

---

If you get no replies, that is probably because everyone absolutely agrees with you.

The most important tool on your side is experience.  There are plenty of times, where the device(s) that you use to collect information malfunctions and you need to make a decision about whether to collect again because of an anomaly or use a separate device to confirm what you are seeing.  You also need to know ballpark numbers that you would expect so you can make adjustments.

Site surveys in general are time consuming.  You don't want to spend all of that time just to collect useless data...

LOL.. we both might be surprised! I actually have a fellow engineer who ONLY does predictive. His reasoning, well Im not buying it entirely. 

I'm sure we all do only predictive surveys until we get caught out there.  Perfectly natural behavior ;)

LOL... just takes getting caught once for me *cough * cough .. 

I also agree 100% on the experience reference. Understanding survey is one of the most important fundamental processes that should be learned by anyone deploying wireless. All to often, its overlooked or not looked at seriously. Knowing the wireless is designed correctlyhelps speed up the troubleshooting process, if you have problems. You dont have to go back and "check" the deisgn. 

I only do active and find predictive to be a waste of time.  The point of conducting a survey is to find the limitations of the structure and then design around those limitatitons or even better to make the limitations work to your favor.

I have a fairly simple set of rules and views on "real" or "virtual" surveying.

1. Virtual surveys are for estimates/tender responses and bid work.

2. Real surveys are for situations where we have an operational or contractual commitment to deliverables to a customer.

That is to say, real surveys are recommended by me where there is a cost consequence of a mistake (specifically to my business, but also the customer obviously). Where it's cost prohibitive or impossible to do a real survey, virtual surveying is fine as long as all parties understand the risks and value of them.

In summary, they both have their place!

Both.  We're on a different model than most with an approving authority in a different geographical area makes the decision on wireless installs.  The predictive surveys are generated to show an estimated coverage area, building layout, equipment/installation requirements, etc..  The post-install passive survey is used to verify the predictive. 

What model would you use in a situation where we have no choice but to send a technician (not experienced in conducting surveys) out to do an active survey (AP on a stick) and relay results back to our engineers to then plan a predictive survey based on the results?

All because of the customer attempting to save costs.

I'd start with a (pessimistic) predictive survey in VisualRF or the like to see where I think the worst trouble spots will be. Then send the tech and ap-on-a-stick to those spots.

Have the tech move the stick and re-test until you can tell how to resolve that trouble spot and move to the next.

Will the tech have internet access, or some other way to connect back to you?  If so, use desktop sharing so you can see the heatmap in whatever tool is on the tech's laptop.

I suppose I should add that we've done this -- since we're in the save money wherever we can camp ;)

Issue we use 3rd party techs who don't necessarily have the right equipment (specturm analyzers) therefore we can only get data from programs such as inSSIDer

That's what we give out techs too!

Thanks. I am trying to come up with a better way to make sure the data we receive and/or process is accurate enough to effectively plan the placement of APs without actually going onsite.