RE: Artificial Immune Systems, Intrusion Detection and Disruption Tolerant Networks
@dana-edwards This isn't a rebuttal but it is a different perspective.
I read the IOTA whitepaper awhile back and had some talks via email with the members of the IOTA team. The concept is nice but there is no solution to doing double spends that doesn't fall down horribly once you follow it to it's logical conclusion. Basically you have to trust that you are getting valid information on the fly from your nearest neighbors because there is no "blockchain" to download and validate. Instead it's a graph. Also I'm not so sure it really qualifies as a directed acylical graph, seemed to me more like a 2 way response request graph. Like a giant distributed mempool.
This allows the system to be fault tolerant and easier to deploy because you don't need a copy of every transaction since the beginning of time. You can just ask your neighbors who ask their neighbors and eventually you get a more complete picture as to the history of that spend.
However it means you have to trust your nearest neighbor despite having no reason to. A malicious attacker could control the entire network with less than 10% of the total nodes. Their answer is to build it into enough devices that it becomes impractical to obtain that necessary 10%.
I wish them the best of luck, but I'm not bagholdin that one
DTNs are already an important part of the world. The internet itself is a DTN. The TCP/IP protocol was the worlds first wide scale implementation of a DTN. The theory being that if the soviets bombed our infrastructure into the stoneage our military would still be able to communicate and mount an appropriate response.
It does this by detecting damage and routing around it. Having been here on the internet for most of it's life, I always find it amusing when people try to reinvent it.
Old grey beard saying, "Those who fail to understand TCP/IP are doomed to re-implement it... poorly".
I do like this topic though. TCP/IP is one example , the old usenet "store and forward" method is another.
I am having a hard time relating your concept of an artificial immune system to the topics though. This seems to me to be just an natural advancement of current security hardening techniques. Or am I missing the point?
If your neighbors are household devices then what is the issue? For an IoT it could work because you don't really need a blockchain for your devices, machine to machine. Of course I could be wrong.
As far as DTN goes, the DTN is a more resilient way of doing things. TCP/IP does have some properties for that built into the protocol but we are talking about IoT and TCP/IP alone is not good enough for moving devices.
Artificial immune systems relate not only to this topic but also to Steemit. If you look at the bot issue then you might be able to solve it by creating an artificial immune system which functions to keep the entire ecosystem of bots in a sort of delicate balance. It relates to one of your own posts about bots.
And a bot is just at best an autonomous agent. So when we are talking about autonomous agents we could be talking about bots or drones. In the context of an intrusion detection system you can use an artificial immune system to provide a layer of security for an IoT, but I'll have to provide more details about how that is possible in a future post.
If you want to investigate on your own, consider looking into how an artificial immune system can benefit Steemit in an environment with swarms of bots, and consider the vulnerabilities an IoT network could have in the context of a smart home for example and how an artificial immune system could function to detect intrusions, or hostile changes.
Please promise me we aren't ushering in this...
What I think of when I hear "internet of things"!
I need sleep, been up all night on steemit again. Will re-read your stuff and comment again in the morning.