MULTIAGENT NETWORK CONTROL FOR THE GROUP
OF UAVs
Oleg Granichin, Konstantin Amelin
In this paper the adaptive network control for multiple
unmanned aerial vehicles (UAVs) is considered by
two sides as problems of a hardware implementation
and an algorithm choice. We develop two levels network
for interactions. High level is for communication
with group and low level is for the interaction between
UAVs. Physical basements of both interactions are radio
and Wi-Fi channel or Internet. But for high level
communications with ground stations we mainly use
Internet or radio channel and Wi-Fi is used for the low
level. There are no strong tasks for the single UAV and
this principle allows increasing the overall performance
for the group which has a general task. To realize this
approach we suggest three layers control UAV’s system.
For example the common group’s task of monitoring
of wide land is studied in the focus of the maximizing
the flight duration through common usage possibilities
of thermal updrafts. For the precise detection
of the thermal updraft’s center the simultaneous perturbation
stochastic approximation (SPSA) type algorithm
is proposed.