Distributing Bandwidth Costs

If you haven’t yet read Bob Cringely’s latest column you’ll find it to be an interesting vision for content distribution in the future. Diving right into the problem of bandwidth for video distribution over IP Bib comes to the conclusion that peer-to-peer technology will be the savior of the content holders in the end preventing them from needing to purchase obscene amounts of bandwidth.

Interestingly I came to the conclusion about a third of the way through the article that he explores further right towards the end. The way to encourage people to share their bandwidth via P2P is to pay them for it. Bob mentions “Peer Impact” which is a content company and software program that has currently adopted a model similar to this. I would take this one step further and say that there is a business model for some company to create a peer-to-peer network which pays each user on a per-megabyte of transfer they do for the network and then turns around and resells that bandwidth to content creators who need bandwidth. If you think about it this is a much more efficient system than the top-down system currently in place. It also solves one of the biggest problems with widespread BitTorrent, etc. systems which is that people leech from the system and then disconnect before returning their fair share of the bandwidth. If people were being paid to simply run a P2P client in the background, one which was agnostic about what was being transferred (eg. you would not have to intentionally download the content yourself first, it would all be automatic), you wouldn’t have a problem with people keeping the client open because they would see it as a moneymaking opportunity which costs them nothing. The ISPs will probably grumble at first but in the end they would come around because it saves them internet bandwidth if the content can move around within their network instead. What you’re really setting up in this scenario is a content distribution and caching system where people dedicate some of there hard drive towards caching content that is currently in demand and then sharing that content with other subscribers on the network. I think there’s a real business potential in this market so if you capitalize on this idea please remember where you got it.

One other issue Bob mentions is the distribution of live streaming content. He seems to see this as some sort of unresolved problem. While it’s true that most P2P systems don’t handle live content very well there is another solution. For years Cisco has been promoting IP Multicasting as a way to conserve a lot of bandwidth for live streaming broadcasts. Unfortunately, the technology is still not widely implemented. If the major backbone carriers and ISPs made a big push to support multicasting on all their routers content creators would only have to push one stream out to the internet where routers would distribute it globally. It’s really a well thought out solution but more people need to get on board. Strangely enough Cringely doesn’t even mention multicasting which leads me to believe this technology is still very much under the radar even though it has been used on some private networks for years.

Comments are closed.