Microsoft P2P: Better Than BitTorrent?

[dubstylee]

Researchers at Microsoft's computer science lab in Cambridge have developed a peer-to-peer filesharing system that they say overcomes the scheduling problems associated with existing distribution protocols such as Bit Torrent.

The researchers claim download times are between 20-30 per cent faster, using their network coding approach, than on systems that only code at the server, and between 200 and 300 per cent faster than distributing un-encoded information.

Naturally, Microsoft is very keen to stress that this technology should be used for distributing legitimate content. It even put that in italics in the press material.

The basic principle of the system, dubbed Avalanche, is pretty much the same as BitTorrent. Certainly the problem it solves is: a large file needs to be distributed to many people. One server does not have the bandwidth to deal with all that traffic, so you need to find another way of getting the file to everyone who needs it.

If the file is broken up into smaller pieces, these can be distributed among a smaller number of people, who can then share the pieces to make sure they all eventually have the complete file.

The problem with this approach, as anyone who has ever tried to download content on the system - legitimate or otherwise - knows, is that towards the end of a download, any one downloader could have a while to wait for the particular pieces he needs. As the number of receivers increases, scheduling traffic also becomes more complex, and the whole process slows down.

Microsoft Research's approach gets around this by re-encoding all the pieces, so that each one that is shared is actually a linear combination of all the pieces, fed into a particular function. The blocks are then distributed with a tag that describes the parameters it contains.

Once you have downloaded a few of these, you can generate new combinations from the ones you have, and send those out to your peers. Collect enough of these pieces, and you will have enough information to reconstruct the whole file. Even if you don't have all the original pieces distributed by the person who held the original version of the file.

Peers can make use of any new piece, instead of having to wait for specific chunks that are missing. This means no one peer can become a bottle neck, since no piece is more important than any other. It also means overall network traffic is lower, since the same information doesn't have to travel back and forth multiple times.

Read the Research Paper from MS (pdf)

Read the complete article (http://www.zeropaid.com/news/5490/Microsoft+P2P%3A+Better+Than+BitTorrent%3F/)

[Auggie2k]

The last person I want to know I'm downloading illegal files is the snoopmaster themselves microsoft!

[The Hunter]

Uh huh, hey Bill, you wouldnt mind if I sent an image of an operating system to a friend would you? J/K

[Auggie2k]

Uh huh, hey Bill, you wouldnt mind if I sent an image of an operating system to a friend would you? J/K
Lol... nice!

[Kyle06]

Heh yea... I think I'll stay with BT

[crackerjacker]

hmm par 2 files are nice
btw various chunks that are being sent via torrent are in no particular order either
*point is*
i am awesome

[zarquon]

This is one bandwagon doomed to failure!

[bobhss]

Well, hopefully it gives ideas to our beloved Bittorrent creator and he makes "improvements" to his code and we all benefit without actually using anything extra from Microsoft.

[ducttapeBigSexy]

For once, it seems that m$ has actually made something that works pretty nice. Too bad they can't come up with anything new - like every other product they've ever made.

[DrainBamaged]

Great. I'll finally have an even faster way to download the latest MSN Messenger and Windows Media Player. Just what I've always wanted.

[Julian]

hmm par 2 files are nice
btw various chunks that are being sent via torrent are in no particular order either
*point is*
i am awesome
You Suck~! But back on the topic... The downside of this protocol is that the proprietor of the content must upload 120% of the file size initially, the upside is that this makes 20% or 5x as many peers with the blocks you want available at any one time. In short, this is a redundancy thing, and not really a _great_ improvement over BT. In short you can recreate the original data with a 20% higher success rate of entire file availability.

[AussieMatt]

If this system is for a Distributed computinng network like http://www.lxsystems.com/technology.html
then there will be seed servers so 120% from the seed server isnt that bad .

Peer Impact a online music store is using LX systems to power the network Peer Impact also gives uploader credits for transfering thier inventory.

www.peerimact.com

[dragoonballz]

can anyone say the RIAA are in on this? Microsoft will probably leak all the info to the RIAA who has probably hired Microsoft to do this, and tricking all its users to thinking that this is some revolutionary thing to the P2P World!

[Matt]

i doubt that this is a conspiracy. what this is is a way for microsoft to probably release patches (similar to grid computing) and will probably be in longhorn but you won't even see it. that's my take on the whole thing. I highly doubt they would unleash anything that could get them into any more trouble.

[muffenme]

Either that or a way to make lots of money.

[ABC123666]

Can I use it to download longhorn?

[crestfallen]

Quit being paranoid. Nobody really knows anything about the protocol except the same information that was written differently in a few articles.

EDIT: D'oh, unless you read the research paper. Or the summary at the Register. Still.

[Digital Bliss]

You Suck~! But back on the topic... The downside of this protocol is that the proprietor of the content must upload 120% of the file size initially, the upside is that this makes 20% or 5x as many peers with the blocks you want available at any one time. In short, this is a redundancy thing, and not really a _great_ improvement over BT. In short you can recreate the original data with a 20% higher success rate of entire file availability.

You swallow but back on the topic... No one is going to buy into this its microsoft for christ sakes thats like saying to a drug dealer that the dea is making a new drug or somting i hope people arnt that stupid

[mindtorrent]

Big suprise........more Microsoft bashing. Get over it please.

[Christoph]

I would use it,if it is faster as bittorent.why not?only becuase there is a branding of microsoft?uhh..C'mon they said that they will get all people with illegal verisons of XP...uhh...yea no one was catched!I use microsoft as OS why not for my P2P?they are not as bad as people always say

[Gamer8585]

M$ will cave to **AA pressure to "monitor" the network, that is if its not already in league with them *Cough*ESV*Cough*

[fireforce555]

Hmmm, a p2p app from MS. I can see the info about it now.


Min SysReqs

1.5Ghz CPU
512 Minimum Memory
Windows XP

And one month after its release it gets a service pack larger than the install was in the first place LOL

[meyou123]

I can't believe that some of you would consider this! Anyone remember the Loki Torrent website and how it was shut down? Well I would not doubt that Microshaft would be in cahoots with the MPAA and RIAA to make a "honey pot" site and then try to collect the IPs of everyone on there.

Crazy you say? Mabye, but I wouldn't put it past them for a minute...they are ALREADY making it so that you have to have a certificate from the owner of whatever you are downloading or uploading before you could use it in the first place! Why would you want to go through the trouble of using this? I would much rather wait until the technology is copied and put into something I can use and trust. At least trust more than Microshaft anyway.

[AussieMatt]

Hope you Micosoft Bashers are using Linux .........if your using Windows you dont have a valid argument becuse any monitoring or copyright enforcing technology will probably come will probably come from the OS like M$ Longhorn and its trusted computing features .

Some of Microsofts p2p Technology has beeen released as open source in the past and whos to say this technology wont be made open source .Publishing a paper gives p2p developers ideas and a new way to do things .If you dont want to use Avalance technology then just dont use it in the future .Remeber p2p doesnt just mean filesharing and copyright infingement .
http://research.microsoft.com/~antr/Pastry/

[muffenme]

The only way I would this program is

1 It's Free( no cost to download or get files )
2. It has no adware, malware, or spyware
3. It has no restriction on what files I share
4. It runs on Windows 2000/98


They can include
1. Just like MSN messager, it would have genreal ad

[meyou123]

The only way I would this program is

1 It's Free( no cost to download or get files )
2. It has no adware, malware, or spyware
3. It has no restriction on what files I share
4. It runs on Windows 2000/98


They can include
1. Just like MSN messager, it would have genreal ad


Well that automatically counts you out then, because of NUMBER THREE!

"It has no restriction on the files I share" Wrong! You have to have a "certificate" for uploading and downloading, stating that the content owner approves. Now, if you do find something you like and get approval for it, then you can use it. But to me, this is useless! It is not p2p, just a way for you to get files you can get anywhere else on the net in a p2p shell.

[PornoCactus]

wow, Microsoft themselves are producing another program that will somehow screw the public. ::claps hands:: anyone who actually dl's this app might as well toss their machine into the nearest lake. We all know that anyone who believes microsoft would ever develop a program that would help the averge PC user is a complete moron. :]

[Wierdy1024]

I think I've noticed a bit of a problem with MS's idea about the network coding thing. Let me know if anyone has more info on this. Here's an email I sent to the guy doing the research:


I have read your research paper on your website and, while the figures are very impressive, I feel that one very important factor has been left out of your simulation that might make a very big effect in the overall result.

The effect is that on most p2p networks connections are usually not stable between peers, often due to filling the bandwidth fully or peers leaving. This causes many block transfers to fail part way through due to lost connections. In my experience with some network hardware this can be as high as 80% of block transfers that fail to complete.

On most p2p networks which don't use any kind of coding, such as bit torrent, this isn't a problem because block transfers can continue from the point that the last peer left them, and there is only the reconnection overhead to a new peer of wasted data.

Using network coding, if I understood it correctly, every coded part from every peer is different. This makes it impossible or very hard to continue the transfer of a coded part from another peer. It might be possible to continue the transfer of a different part from another peer and then use very complex methods to recreate the original, depending on how the network coding is done, but that is mainly over my head. If it isn't possible then there are huge amounts of data that get transfered that becomes useless, and the whole efficiency of the network would suffer considerably as a result. If all peers had connection loss rates that are similar to current p2p networks (such as ed2k and gnuttella which have this problem worst due to high numbers of concurrent connections) then the overheads caused by connection losses and therefore wasted data transfers would far outweigh the benefit of network coding.

I hope that a workaround can be found!

Thanks

[Afn]

I think this shows how p2p really works.

If you have 1 seeder (a person who has 100% of a program) and 5 downloaders, if you tweek the downloading so downloaders send 20% of the program to another peer before download is completed, then you could solve a leech problem.

The problem is, there are MANY programs where there are 1 to 2 downloaders with fractured downloads and no seeders.

For p2p to work, you need one seeder and 5 downloaders. This could be built into the system. What you share, can not be unshared until it is seeded by 5 people, and so on.

Availablity is the bigest problem on p2p.

[davehutt]

yeah sure its p2p but i bet it ain't free and if it waz microsoft wud sell you yanks off to RIAA ASAP,im from UK and hardly any one gets sued(only 30 ppl or sumthin eva),so i can get away with twice as much as u lot but id luv 2 b in Canada, lucky buggers. still i think if some1 cud clone the idea and make it freeware it wudnt b so bad but microsoft in p2p is a no go.

[Wierdy1024]

IT's good technology they use and ought to stop the so many dead download problems but the DRM and security measures will stop "Freedom of speech" for people to share what they like - so it's simple: we wait till MS makes the avalanche app and then make our own compatible client to the network sop we can anonymously use their network for our uses - remember it's a trackerless network so it ought to be possible to use it anonymously.