FIELD: wireless communication.
SUBSTANCE: route detection procedure is executed to determine the route from the source node to the recipient node, wherein the reinforcement learning procedure is performed in the route detection procedure, wherein the source node performs broadcast transmission of a route request message wherein a repeat broadcast transmission of a certain number of segments is performed with respect thereto by one or several nodes receiving the route request message, wherein the recipient node performs broadcast transmission of a route response message in response to receiving said route request message wherein a repeat broadcast transmission of a certain number of segments is performed with respect thereto by one or more nodes receiving the route response message, wherein the node that received said route request message or said route response message updates the routing table with the corresponding estimated reward values for a corresponding action, the reward function is used in the route detection procedure in order to estimate said reward value, dependent on the indicator of the strength of the received signal measured during the reception of said route request message or route response message and on the amount of segments distributed in the network through the route request message or route response message in order to reach the receiving node.
EFFECT: providing a possibility of initialising the routing table configured in the node for making a decision on the next segment (hop) when routing packets in the direction to the recipient.
14 cl, 16 dwg
