Signal Boost and Speed

[Aggie79-82a]

I did not want to hijack the recent signal boost thread with this question.

 

Does boosting the signal from the RV Park WIFI provide any increase in internet speed? I find I can connect but depending on which park antenna I use, the speed varies from real slow to almost okay. Willing to invest in signal booster if it helps on the speed (upload/download).

 

 

[Mark and Dale Bruss]

You can only go up to the speed of the Park Internet feed. If that is slow, then you will be slow no matter what speed you connect to the Park with. Also the number of concurrent users in the Park can affect your speed.

 

But a poor connection to the Park can slow you down more than the Park's speed.

[TCW]

Every time the Wifi signal is repeated, there will be a reduction in the speed of the connection, possibly as great as 50%. A single radio repeater will result in a greater speed loss than a two radio repeater, since it can not transmit and receive at the same time. I have found that in parks/campgrounds that use repeaters rather than have each access point connected directly to an internet source, connecting to the access point closest to the primary access point often results in the best speed even if it is not the strongest signal available. The more hops you can eliminate while still having a usable connection, the higher the speed you will experience. In my experience, a booster(wifi adaptor) that is connected directly to the computer will provide a higher speed than one that creates a hotspot and is connected to by wifi if everything else is equal. For example: When we stay at a relative's house that has Comcast internet, direct connection to the router/modem givens download speeds of about 90MBS. Connecting wirelessly gives speeds in the 40-50MBS range. Connecting wirelessly through my WFR Go2 results in speeds of 25-30MBS. At those speeds it is not a real big deal for what I do on the computer, but when using a park or public wifi system, the affect can be quite noticeable.

[SpaceNorman]

...When we stay at a relative's house that has Comcast internet, direct connection to the router/modem givens download speeds of about 90MBS. ...

 

I'm curious how you've ascertained a download speed of 90Mbs? According to the Comcast web site - the fastest package offered is their "Blast" package which advertises speeds up to 75 Mb. Typically maximum speed (thruput) is controlled via configuration parameters applied to the upstream router port and can't be exceeded.

[Mark and Dale Bruss]

http://www.speedtest.net/

[budeneighe]

Personally, I prefer www.speedof.me over Speedtest.net. Both do provide some graphical clues about how consistent the throughput is and that can be more of the bogginess one may feel at the keyboard than the average speeds. If a speed test shows a speed of 4mbps but takes 10 minutes to complete, one should be suspect of the accuracy of the testing methodology.

 

One other often overlooked factor is what level of service is the park paying for. Fewer guarantees, such as a set minimum throughput guarantee, will cost them more. Another is the service that they are tapped into. In this case, being a part of a service that also supplies business services for a nearby shopping center can really slow throughput when the businesses are open and working.

 

People usually assume that the delivered rate to the campground is constant but that is almost never the case. Just as within the park, what that incoming service is shared with can have more affect on the quality of service at the RV than any other factor. It is unlikely without a lot of monitoring that an end user will be able to tell whether their speeds are being more impacted by activity in the park or outside of it.

 

Same is true for cell delivered services. One can have blistering speed to the tower but the trunk services that feed the local area can be bogging down just due to the sheer number of people in that town or area and they don't even have to be directly using wireless data. They can be just a lot of people tweeting and using their cell phones.

 

Jumping to the conclusion that end user performance will be improved dramatically by throwing $$$ at boosting the local signal may not be satisfying in the end. For example, I have gotten 12mbps download speeds over a -104dbm signal and boosting it to -85dbm made no difference in this performance. If the problem was really about signal strength there should have been a dramatic improvement unless I was already at the limit of the backhaul or trunking services. You can't get more water out of a firehose than is available at the hydrant.

[SpaceNorman]

I earn my living as a network analyst for a large financial institution - and spend my days dissecting communications traversing my company's WAN (wide area network) - in many case at the packet level. Every day, I'm engaged to examine reports of "slow performance". In all the years I've been doing this - one universal truth has emerged. That truth is this: Rarely, can the root cause of reported "slow performance" be tied back to actual network problems. Occasionally, network capacity issues along a given link (i.e., more traffic across a given link than the link is capable of carrying) are at the root of the problem. However, for every one instance where network capacity is the issue - there are ten plus instances where the root of the problem is an issue in the application infrastructure (i.e., server misconfiguration, application level interaction between the various components in a multi-tiered platform architecture (i.e., web servers, app servers, database servers, shared storage (SAN) devices, load balancers, etc.) In many cases, it's plain ol' piss poor application design - that doom an application to be a poor performer before the first user ever logs onto it.

 

My point is this. Everybody is quick to point to the network infrastructure as the reason that the user's "click to glass" experience (i.e., the wait time between when a user pushes the button and when the desired results appear on the screen) falls way short of what they desire - when in fact, the real problems sit elsewhere in the end to end communications.

[TCW]

...I'm curious how you've ascertained a download speed of 90Mbs?...

The point I was trying to make was about the reduction in speed when using a repeater/wireless connection not the actual speed, but here you go. I don't have the Go2 running at the moment, but I think this supports what I was saying well enough.

Direct connection to router by ethernet.

 

Wireless connection.

[Jack Mayer]

All the above is true. Especially note what SpaceNorman said.

 

The bottom line is that RV parks have great challenges in supporting their clients, and there is no magic solution. Having been involved at the design and implementation level in RV parks for at least 12 years I can pretty much tell you all the issues. And share lots of sob stories with you.

 

BTW, we DO HAVE a guaranteed 100 mbps fiber backhaul on our system. It does maintain at least 92 mbps. I've never seen less than that hardwired. Even with that - on an unrestricted network I cannot maintain a reasonable user experience. I HAVE to restrict streaming.