Jump to content


Check out our Community Blogs

Register and join over 40,000 other developers!


Recent Status Updates

View All Updates

Photo
- - - - -

bandwidth


  • Please log in to reply
12 replies to this topic

#1 jwxie518

jwxie518

    Speaks fluent binary

  • Senior Member
  • PipPipPipPipPipPip
  • 517 posts

Posted 29 June 2011 - 08:27 AM

Till this date I still don't understand what bandwidth is in Internet network. I know by definition it is the amount of data that can be transmitted through the channels during a specific period of time.

Faster services usually comes with higher bandwidth. As a Cable user in New York City, the company oversold (due to the its monopolistic operation), and so hundreds, if not, thousands of users are usually assigned to the same pool.

Does higher bandwidth really help? The dilemma is that after downloading a couple GB files, or even just a few MB, the Internet speed becomes slower. What usually happen is that after using the Internet for a few days the internet is as slow as snail for the next few days (sometime up to a week). Is that due to bandwidth cap? Is there even such thing as "bandwidth cap" today?

Does more bandwidth help?
15Mbps down / 2Mbps up

This doesn't look good because upstream is only 1/7 of the downstream.
  • 0

#2 WingedPanther73

WingedPanther73

    A spammer's worst nightmare

  • Moderator
  • 17757 posts
  • Location:Upstate, South Carolina
  • Programming Language:C, C++, PL/SQL, Delphi/Object Pascal, Pascal, Transact-SQL, Others
  • Learning:Java, C#, PHP, JavaScript, Lisp, Fortran, Haskell, Others

Posted 29 June 2011 - 08:44 AM

The other thing that many ISP's do is throttling. If you are using a large percentage of the pool, they'll just cripple your bandwidth temporarily so it doesn't affect other people's experience.
  • 1

Programming is a branch of mathematics.
My CodeCall Blog | My Personal Blog

My MineCraft server site: http://banishedwings.enjin.com/


#3 gregwarner

gregwarner

    Obi Wan of Programming

  • Expert Member
  • PipPipPipPipPipPipPip
  • 1586 posts
  • Location:Arkansas
  • Programming Language:C, Java, C++, C#, PHP, Transact-SQL

Posted 29 June 2011 - 08:49 AM

Let me address your questions in reverse order.

The drastic difference between upstream and downstream isn't usually a big deal to end users, since 99% of what they're going to be doing (I just made that percentile up) is downloading. Your average end user won't be uploading much, unless they're running a server or sharing lots of files on a P2P network.

There is a such thing as a bandwidth cap with some ISP's, but you're going to have to check the fine print on your service contract for your particular ISP to determine whether you are being capped or not. Some ISP's slow your connection down when you exceed the cap (EDIT: WingedPanther beat me to the post. This is what he is referring to with "throttling"), some cut you off entirely, while still others do nothing or even charge you per megabyte or gigabyte exceeded. You'll have to look into what your ISP does if you are concerned that you may be exceeding your cap.

Higher bandwidth does help, but you must remember that your connection speed will only be as fast as the slowest link in the chain. (A chain is only as strong as its weakest link.) If you are trying to download from a server that is connected to the internet on a T1 line (1.5 Mbps), you're only going to download at 1.5 Mbps, even though your connection can go as fast as 15 Mbps.

Also, congestion is an issue, and as you have noticed, during certain points in the day, you notice your connection speed slows down. This could be due to more users on the cable company's network at that particular time, or congestion on the other end of your transmission. You can call your cable company when you're experiencing the slow speeds and issue a complaint, and you might get some details as to what's going on that way.

But to sum it all up, yes, bandwidth (or more accurately "throughput") is the amount of data that can be transmitted during a specified period of time.
  • 1

ti-99-sig.png
Hofstadter's Law: It always takes longer than you expect, even when you take into account Hofstadter's Law.
– Douglas Hofstadter, Gödel, Escher, Bach: An Eternal Golden Braid


#4 Alexander

Alexander

    YOL9

  • Moderator
  • 3963 posts
  • Location:Vancouver, Eh! Cleverness: 200
  • Programming Language:C, C++, PHP, Assembly

Posted 29 June 2011 - 11:21 AM

If you are not over the bandwidth cap I would suggest contacting your ISP for diagnostic purposes, assuming the speed drops are not common in your area.

Also a tidbit of information, bandwidth speeds are often listed in megabits rather than bytes, so your actual speeds are eight times less (your upload is 250kBs which seems fine for general internet user, document upload etc.)
  • 1

All new problems require investigation, and so if errors are problems, try to learn as much as you can and report back.


#5 gregwarner

gregwarner

    Obi Wan of Programming

  • Expert Member
  • PipPipPipPipPipPipPip
  • 1586 posts
  • Location:Arkansas
  • Programming Language:C, Java, C++, C#, PHP, Transact-SQL

Posted 29 June 2011 - 11:30 AM

This is probably a little more information than what the OP is asking for, but I thought I'd post it in case anybody is interested.

The term "bandwidth" comes from the terminology used in signal transmission, and when used in reference to radio and copper line transmissions, it refers to the literal width of the band of frequencies used to transmit the signal. It didn't have anything to do with the rate at which data could be sent, although one could calculate the data rate (throughput) based on the known bandwidth, symbol set, data encoding protocol, noise ratio, etc. Bandwidth was measured in Hertz (or KHz or MHz, depending on the scale) and throughput was measured in bits per second.

Now days, the term "bandwidth" has really lost its relevance in terms of actual band widths, as we are no longer concerned with the carrier medium, but rather the practical usefulness, i.e., how much data we can move.

Just FYI, since I find that interesting.
  • 0

ti-99-sig.png
Hofstadter's Law: It always takes longer than you expect, even when you take into account Hofstadter's Law.
– Douglas Hofstadter, Gödel, Escher, Bach: An Eternal Golden Braid


#6 Alexander

Alexander

    YOL9

  • Moderator
  • 3963 posts
  • Location:Vancouver, Eh! Cleverness: 200
  • Programming Language:C, C++, PHP, Assembly

Posted 29 June 2011 - 11:34 AM

@Greg, I thought there was a need for clarification on the term as well. You have hit spot on.
  • 1

All new problems require investigation, and so if errors are problems, try to learn as much as you can and report back.


#7 jwxie518

jwxie518

    Speaks fluent binary

  • Senior Member
  • PipPipPipPipPipPip
  • 517 posts

Posted 30 June 2011 - 06:17 PM

Thank you to all three of you. Now I understand.

Not really. One cable for 3 floors, 6 computers. Not cool -_-
I am still waiting for FiOS.
  • 0

#8 fkl

fkl

    CC Devotee

  • Senior Member
  • PipPipPipPipPipPip
  • 417 posts

Posted 02 July 2011 - 05:39 AM

Another factor relevant is the concept of "net neutrality" (Google it). But the idea says entire traffic over the internet should be free from any sort of priority or prejudice.

I work for a company which provides alternate means so i am not giving my personal opinion, rather just explaining the service provider (ISP) point of view.

No matter how big bandwidth pipes they obtain it will always be limited and the application consumption rate is also growing as the capability of medium. So there is always a scenario in which an average user running torrents or streaming or whatever, which slows down a more critical user's bandwidth for e.g. sending emails or some other high priority data.

To prevent this companies often like to obtain services that operate on layer 7 called "Deep Packet Inspection" (DPI). There is at least one open source implementation (ipoq) available. What they do is, that they go into application layer of every packet and try to determine which application is it running i.e. bittorent, skype, streaming or what not. Some thing resembling wireshark but far more sophisticated. Then then provide the ability to control (drop / police) etc. such traffic. Also they provide means to quantify which user played how much traffic of what type and charge them accordingly.

I am not arguing that this is "fair". But to me one person bringing down every other person's internet by downloading a pirated video should be worthy of this treatment.
  • 0
Today is the first day of the rest of my life

#9 WingedPanther73

WingedPanther73

    A spammer's worst nightmare

  • Moderator
  • 17757 posts
  • Location:Upstate, South Carolina
  • Programming Language:C, C++, PL/SQL, Delphi/Object Pascal, Pascal, Transact-SQL, Others
  • Learning:Java, C#, PHP, JavaScript, Lisp, Fortran, Haskell, Others

Posted 02 July 2011 - 06:14 AM

Another factor relevant is the concept of "net neutrality" (Google it). But the idea says entire traffic over the internet should be free from any sort of priority or prejudice.

I work for a company which provides alternate means so i am not giving my personal opinion, rather just explaining the service provider (ISP) point of view.

No matter how big bandwidth pipes they obtain it will always be limited and the application consumption rate is also growing as the capability of medium. So there is always a scenario in which an average user running torrents or streaming or whatever, which slows down a more critical user's bandwidth for e.g. sending emails or some other high priority data.

To prevent this companies often like to obtain services that operate on layer 7 called "Deep Packet Inspection" (DPI). There is at least one open source implementation (ipoq) available. What they do is, that they go into application layer of every packet and try to determine which application is it running i.e. bittorent, skype, streaming or what not. Some thing resembling wireshark but far more sophisticated. Then then provide the ability to control (drop / police) etc. such traffic. Also they provide means to quantify which user played how much traffic of what type and charge them accordingly.

I am not arguing that this is "fair". But to me one person bringing down every other person's internet by downloading a pirated video should be worthy of this treatment.

I think a good way to express the fairness aspect: Let's say two customers are sharing a 4GiB pipe, and each has purchased 3GiB service. Now, as long as they don't both attempt to max it out at the same time, there's no problem. However, if Customer A starts downloading movie torrents 24/7, that will pretty well max out the 3GiB that person paid for, only leaving 1GiB for the other person. The provider has three options: expand the pipe to 6GiB, let Customer B deal with the 1GiB that's left, or throttle Customer A to avoid a negative impact on Customer B's purchase. Since most plans in the US advertise "Up to XX GB/s" anyway, throttling Customer A for abusing the bandwidth at the expense of another customer starts to become appropriate. Another option might be to simply prioritize Customer B's traffic, so that Customer A gets as much as is available, but can't interfere with Customer B.

Which option is fair? Expanding the pipe incurs a cost on the ISP beyond what should be required for normal use. Letting Customer A hog the pipe is unfair to Customer B, who is now getting 1/3 the bandwidth paid for. Throttling Customer A or reducing the priority of Customer A's traffic is a denial of Customer A of bandwidth purchased. At some point, someone has to pay, but who's the one being unreasonable?
  • 0

Programming is a branch of mathematics.
My CodeCall Blog | My Personal Blog

My MineCraft server site: http://banishedwings.enjin.com/


#10 fkl

fkl

    CC Devotee

  • Senior Member
  • PipPipPipPipPipPip
  • 417 posts

Posted 02 July 2011 - 06:31 AM

Yeah conceptually i agree WP and that is one of the reasons i am fine with working in this domain.

However, the counter arguments given by every advocate of Net neutrality (and believe me that is a large volume of global highly effective ones) are:

Internet should not be discriminated based upon the identity of the sender / receiver. So you cannot distinguish based upon customer A and B in your relation above.

Reading L7 brings many complicated issues such as loss of privacy although it is explicitly elaborated that L7 is used to identify the type of traffic i.e. it is restricted only to knowing if it is running streaming, skype torrents or what.

People go up to saying that making any impact over the routing of a packet depending upon who was the sender is entirely against the whole spirit of internet.
  • 0
Today is the first day of the rest of my life

#11 WingedPanther73

WingedPanther73

    A spammer's worst nightmare

  • Moderator
  • 17757 posts
  • Location:Upstate, South Carolina
  • Programming Language:C, C++, PL/SQL, Delphi/Object Pascal, Pascal, Transact-SQL, Others
  • Learning:Java, C#, PHP, JavaScript, Lisp, Fortran, Haskell, Others

Posted 02 July 2011 - 08:26 AM

I suspect that more and more of the traffic being sent will be encrypted to deal with privacy issues. I also think there are a lot of people from my era, where the internet existed primarily as a medium for communication between the various universities and government agencies that had spawned it. The reality, today, is that the Internet is a commercial enterprise.
  • 0

Programming is a branch of mathematics.
My CodeCall Blog | My Personal Blog

My MineCraft server site: http://banishedwings.enjin.com/


#12 fkl

fkl

    CC Devotee

  • Senior Member
  • PipPipPipPipPipPip
  • 417 posts

Posted 02 July 2011 - 11:24 AM

Yes Indeed! Encryption is getting more and more common, but so are DPI techniques. There are a good deal of research papers on how SKYPE obfuscates itself to get through firewalls and NATs. Still there are all sorts of mechanisms ranging from deep mathematics like entropy calculation, traffic behavioral patterns etc. to more programming intensive ones like knowing torrents makes a lot of connections simultaneously in the start some of which might contain some tracker communication. So sometimes they try sending spoofed torrent requests on tracker ip's i.e. there is always some initial part of encryption (Key negotiation etc) which is unencrypted. If they simply get a rejected response, it confirms the app is some torrent.

Remember the objective is not to decrypt the communication (which is the actual hard part), but some what lesser in scope i.e. we only need to determine which app is it. So there are many kinds of techniques which will keep on expanding and i am sure if people are criticizing reading application header of ordinary packets now, they would be questioning these techniques too some time latter.

I only referred to app header thing so that audience not related to network security can have a rough understanding of how DPI works.
  • 0
Today is the first day of the rest of my life




Recommended from our users: Dynamic Network Monitoring from WhatsUp Gold from IPSwitch. Free Download