Self-assembly for discreet, fault-tolerant, and scalable computation on internet-sized distributed networks

摘要: When engineers compare biological and software systems, the former come out ahead in majority of dimensions. For example, human body is far more complex, better suited to deal with faulty components, resistant malicious agents such as viruses, adaptive environmental changes than your favorite operating system. Thus it follows that we, engineers, may be able build systems ones we today by borrowing technologies from nature injecting them into our system design process. In this dissertation, I present an architectural style accompanying implementation support for building distributed allow large networks, Internet, solve computationally intensive problems. This style, tile based on a nature's crystal growth, thus inherits some dependability, fault adversary tolerance, scalability, security. The allows one distribute computation onto network way guarantees unless someone controls fraction network, they cannot learn private data within or force fail. These are highly scalable, capable dealing nodes, discreet since every sufficiently small group nodes knows neither problem nor data. The formal mathematical model self-assembly. In order leverage software, define notion self-assembling develop compute functions adding, multiplying, factoring, solving NP-complete problems SubsetSum SAT. each system, prove its correctness, probability successful computation, show running time tileset size asymptotically optimal. I use assembly analysis proving built using discreet, fault- adversary-tolerant, scalable. further implement tile-style empirically evaluate style's utility.