Method for Intelligent Routing Within Ad-Hoc Networks with Complex Topology

The most common problem for all sensory IOT networks is routing inefficiency. “Effective” here means that wandering packets shouldn't pass excessive or overloaded hops as well as shouldn't generate extra official traffic to coordinator. These are critical conditions in low-end networks, such as networks, formed of sensors, connected both together and to the internet. This work proposes solution for “effective routing” problem by use of machine learning through genetic algorithms. This reduces both computational time and energy draining in all weak hops through network. But here comes out the first valuable weakness: stable coordinator with an ability to simulate neural network, i.e. to perform complex parallel computations, is required.