Swarm
Intelligence Swarm
Intelligence is a design framework based on social insect behavior.
Social insects such as ants, bees, and wasps are unique in the way
these simple individuals cooperate to accomplish complex, difficult
tasks. This cooperation is distributed among the entire population,
without any centralized control. Each individual simply follows
a small set of rules influenced by locally available information.
This emergent behavior results in great achievements that no single
member could complete by themselves. Additional properties swarm
intelligent systems possess include: robustness against individual
misbehavior or loss, the flexibility to change quickly in a dynamic
environment, and an inherent parallelism or distributed action.
There
are four principles governing the swarm intelligence. These are:
Positive Feedback - reinforces good solutions present in the system
Negative Feedback - removes old or poor solutions
Randomness - so solutions can be tested regardless of perceived
quality, which in turn, result in creative and unconventional
solutions
Multiple Interactions - key to building up the best solutions
By
understanding these properties and applying them correctly, swarm
intelligent systems may be designed. Each principle plays a clear
role in governing the system.
The
benefits of swarm intelligence can work effectively to resolve current
issues in MANETs, or mobile ad-hoc networks. A MANET is a collection
of computers, or nodes, participating and cooperating in a computer
network. MANETs are increasingly appearing now that wireless devices
become more and more ubiquitous. Information is communicated between
nodes via a wireless link. There is a limited communications range
for each node, and each node has only a few neighbors. Neighbors
are nodes that can communicate directly. Nodes are assumed to be
mobile; nodes can move relative to each other. Mobility causes the
topology of the network to be quite dynamic.
Self-Optimizing
Auction Systems
Bluetronix's
new protocols are being developed to account for the dynamic topology
of ad-hoc networks, rather than try to adapt old approaches to new
problems. Ad-hoc networks pose a fundamentally different set of
issues than traditional wired networks. Swarm intelligence is a
novel approach which can account for a larger set of critical metrics,
as well as to adapt to highly variable factors such as network size
or node speed. It must also be mentioned that the very metrics to
which the network must adapt are also subject to rapid change.
The
principles of Swarm Intelligence can be applied to a variety of
other applications reaching far beyond computer networks. Where
a centralized design fails, Swarm's unique collective (or distributed)
problem solving method may prove an attractive alternative. For
example, traffic congestion is often reduced by increasing the number
of lanes. This does not always work. Using a swarm paradigm to model
for traffic patterns, making the road longer and manipulating the
speed limits has been shown to reduce gridlock and actually decrease
travel time in certain cases.
Optimizing
scheduling or distribution tasks can be very time consuming, or
even virtually impossible in some instances. Southwest Airlines
has used swarm to develop a more efficient model of cargo handling,
saving the company $2 million per year in labor costs. General Motors
Corp. implemented software using adaptive technology to schedule
car paint jobs and to avoid the scheduling conflicts from which
the manual system suffered. The new system resulted in a 30% productivity
improvement and 35% fewer business-process changes.
