Both repositioning robots, as well as robot
node communication, requires energy expenditure. The energy expenditure
by robot motion can in some cases be a lot more significant than the one
required by communication.
??? Combining communication constraints, sensor fusion requirements, energy
constraints, navigation limitations, and adaptive sampling into a distributed,
scalable deployment algorithm. A scalable way to combine the
results of communication, navigation, and information optimization problems
is via the potential fields method. MWSN nodes move according to
38
Chapter 2 Algorithms for Robotic Deployment of WSN
forces calculated from network, navigation and information ???energy-like???
functions.
We can formulate the network model of the mobile sensor nodes based on
results from congestion control. The flow control problem through an ad-hoc
sensor network can be separated into two problems:
??? The ???routing problem??? is similar to a ???ravelling salesman??? NP-hard problem
in that it aims to select data routes (???hops???) with minimal cost between
the wireless nodes forming a graph. The ???cost??? can be defined in
terms of geographical distance, energy consumed, or time delay through
the network.
??? The ???congestion control problem??? consists of finding and regulating the
optimal flow rates between the network nodes in the presence of network
capacity constraints. This problem can be further decomposed into a static
optimization problem (finding the optimal flow) and a dynamic stabilization
problem (converging to the optimal flow)
entiators between di?®erent network protocols, but we will assume that the
communication hardware provided to the robot nodes has been a-priori chosen.
Pages:
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101