Network Laws

From 118wiki

Jump to: navigation, search

Network and Computing Laws

The following "laws" are not like Laws of Nature (Newton's Laws, the Laws of Thermodynamics, etc), but rather they are patterns of our current technological advance. Some of these patterns are justified by enduring principles, and others may not hold up over the years if technology shifts radically.

These laws are not often covered in traditional network courses; they are valuable to know even if you are not a networking specialist. Many "information technology" positions in large organizations emphasize general business knowledge, if not knowledge pertaining to the specific organization or its business sector, at least as much as technical expertise (for instance, this article shows some data about this). These laws are part of the present "culture" of network commerce and you should try to become literate about such things, even outside of this networks course.

Moore's Law 
Wikipedia defines Moores Law as the market-place phenomenon that complexity of integrated circuits doubles every two years. This has usually been attributed to manufacturing small transistors, but sometimes there are other techniques.
Benford's Law 
Wikipedia has a good explanation of Benford. (Benford's Law doesn't really have anything to do with networks, however you should learn about it sometime.)
Gilder's Law 
is the observation that the total network carrying capacity is doubling faster than processing power doubles; there are several references to this so-called law (but few authoritative studies).
  • In this article is a big list of "technology laws", but I don't ask that you know all of those.
Metcalfe's Law 
Originally, Metcalfe's Law stated the value of a network grows quadratically with the number of users. This may be overstated or understated (an extreme form of this is Reed's Law). A recent article in IEEE Spectrum declares Metcalfe's Law is Wrong (and you should read at least enough of this to understand the argument).
Network Effect 
The Network effects is closely related to, but less quantified than, Metcalfe's Law. The wikipedia entry on the network effect is well worth reading; a closely related idea is the Tipping Point (follow the external link in the wikipedia article to find a concise summary).
Small-World Phenomenon 
The small world phenomenon is an empirical observation about the structure of natural-world networks as they scale (please read about scalability). A gentle introduction to the mathematics of the small world phenomenon is this article by Jon Kleinberg. An example of the phenomenon in a study of an online community is described in this article. A more detailed look at the mathematics is found here.
Power Law Distribution 
Many of the aspects of network content, such as web pages, web sites, have been studied and the popularity measured. The outcome of many studies is that the popularity frequency distribution satisfies a Power law distribution, such as Zipf distribution.
Long Tail 
Power law distributions are typically long tail distributions, and this fact has important consequences on network behavior, how commerce benefits from networks.
Godwin's Law 
(NOT a real law, but funny). Godwin's Law -- read only if you are curious.
Personal tools