Question

Locked

How do we verify HAN technology throughput?

By AYChen ·
The big numbers that HAN transport technologies promise these days are more than mind boggling. They often do not have any real-life data to support the claims. A concise procedure that an ordinary user may be able to duplicate is needed to close the gap. Do you know any procedure that can deliver such? For example, how about publish performance numbers based on file transfer between two PCs that are interconnected via two HomePlug AVx adapters? If PC CPU speed and Ethernet interface may affect the performance, how about specify what characteristics are required? We can even start with the two adapters are on the same duplex jack as the reference point.

This conversation is currently closed to new comments.

10 total posts (Page 1 of 1)  
| Thread display: Collapse - | Expand +

All Answers

Collapse -

Your post is as vague as the purported claims you are questioning.

by seanferd In reply to How do we verify HAN tech ...

How about listing specific claims, and links to where said claims are made?

On the flip side, there is no obligation for any company to provide specs on all hardware and cabling you might use - knowing the limits of your equipment is your job.


You, of course, are free to test raw throughput on such a device, which is all an industry claim can possibly cover.

edit: Now, if claims are made that throughput will be such and such regardless of environment, such claims may be invalidated on their face.

Collapse -

I am an engineer trained with traditional engineering disciplines

by AYChen In reply to Your post is as vague as ...

It is evident that you are among the new generation who is very much used to the contemporary marketing tactic of "blame the victim". Too much of the performance numbers are derived from "ideal" or even "theoretical" conditions. If a product is intended to be used in an everyday environment, performances should be referenced to such. So that, an end user can have some idea what is reasonably possible, before jumping into the frustrations. Leaving consumer to play cat catches mice game by selling products to them first is purely irresponsible and wasteful.
To be honest with you, I have found that actual throughput numbers of most HAN transport technologies are only a fraction of what they are stated as "up to" values. Yet, when I challenged the vendors, they often came back with a compliment asking me how did I get such "good" numbers, because they own are even much worse. Of course, on an open forum like this, I shall refrain from naming the actual identities of such parties.

Collapse -

So to clarify

by robo_dev In reply to I am an engineer trained ...

You are looking at LAN benchmark applications and techniques?

Passmark makes an excellent product for that:
http://www.passmark.com/products/pt.htm

Their network benchmark test is quite good:
http://www.passmark.com/products/pt_advnet.htm

Collapse -

Let's Try Mom and Pop Friendly Approaches

by AYChen In reply to So to clarify

Thanks for your URL's. However, they are going toward the opposite direction as I was driving at. That is, to classify and claim a product "home ready", we need to present performance numbers that can be verified by using tools suited for true consumers, not a techy.

Collapse -

That's a very difficult assignment

by robo_dev In reply to Let's Try Mom and Pop Fri ...

I hate to be dismissive of your request but,
most 'non-tech' users I know only have a vague understanding of exactly how the Internet gets to their computers, much less care about making sure it's the fastest method.

Most users I know are not at all interested in how it works, they simply want it to work, and if it is not fast enough, they would call someone to make it work faster, and not attempt to install and run a utility that they do not understand.

Further, any testing tool requires a great deal of technical knowledge, as the user would need to know, for example, what interface they are testing, what the network addresses of the relevant interfaces are, etc, etc. Even interpreting the results is problematic, since a non-technical user may not understand that 200ms is a slow ping time versus 2ms, and that 10MBs is not particularly fast....although it may be perfectly adequate if they are merely surfing the Internet and not copying DiVX files over their LAN.

Collapse -

Let's go back to basics

by AYChen In reply to That's a very difficult a ...

Your statements are all factual. However, this knowledge / skill chasm is the result of how the current digital age is developed and evolved. Although marketing people are responsible for this disparity, technical people should bear much of the blame because the hold-back from educating the mass.

Before I get into the solution or answer to your challenges that are based on detailed technical considerations, I like to start from reviewing conventional communications. That is, how do consumers get involved with troubleshooting traditional telephone services? Are you aware of the procedure for any subscriber to divide the responsibility of a trouble situation between the telephone company and themselves? (Hint: It is printed in the "Before Calling Us" section of most white-pages published by your local telephone company.) Once we have a common reference point, I will quantify the broadband task that I posted originally in a few steps to demonstrate how could Mom and Pop be empowered to be their own IT manager. It will not solve all the problems. However, it will carve out the basic but most critical networking issues for Mom and Pop, leaving the tougher ones for the professionals.

Collapse -

Just to clarify

by robo_dev In reply to Let's go back to basics

I am here to help, and if you ask my opinion, I will offer it.

What is it you are trying to accomplish, exactly?

Are you trying to promote the sales of wireline networking products? If so, that's great.

Are you looking to buy or develop a utility to validate your product performance claims?

If I understand you correctly, you are assuming that both small business and home users would use some sort of measurement tools to help them to select the best network equipment to use?? My opinion is that they would not use such a tool, and would not base their buying decisions on such a tool.

Collapse -

You lost me on the first sentence

by robo_dev In reply to How do we verify HAN tech ...

The term "HAN" is used for several different things, so I assume you are referring to Powerline Networking, as defined by the HomePlug standards?

HomePlug devices can provide 85-200 MBS over home power wires, depending on the device. This requires a $100-200 gizmo which won't work if the power goes off. You cannot plug a powerline adapter into a UPS, as the signal will be blocked.

I don't consider this to be mind boggling since the CAT6 wiring in my house can provide Gigabit Ethernet to each device at much lower cost.

Most HomePlug devices provide "up to" 85 MBS or "up to 200 MBS" of throughput. Is your goal to develop standards to determine what the real-world throughput of these devices is?

Collapse -

A bit of clarification

by AYChen In reply to You lost me on the first ...

My apologies. I was making comments to the HomePlug AV2 article. So, my assumption was that HAN (Home Area Network) would be relatively clear expression to readers interested in this subject. HAN covers the whole general topics about networking around a home, not just HomePlug x, but also HomePNA x, WiFi x (802.11x), etc., plus other issues and topics involved in setting up a working network.

Yes, one must have the basic understanding that HomePlug x devices are to be plugged into electric wall outlets directly. Otherwise, surge protection circuitry in UPS and certain type of power strip will attenuate HomePlug signals.

If you put a basic criterion of safety into setting up a HAN, you will know that physical Ethernet based on CAT5 and up cabling is NOT an viable deployment approach! However, the above mentioned three categories are, because they are based on either no-wire or no-new-wire media. But, do not get to up-tight about my statement. Ethernet basics are in all of these three categories anyway. :-))

I am not asking for any technical standard. I am just reminding everyone that there was good old business ethics that we appear to allow to disappear. That is, just be a bit straightforward, by giving consumers something they can put their hands around, instead of jumping around, pointing fingers or find excuses to avoid facing the truth.

Back to Software Forum
10 total posts (Page 1 of 1)  

Related Discussions

Related Forums