Simulation-Based Application of Flying Ad-Hoc Networks with Mac Layer Adaptation

Abstract

Flying ad-hoc networks (FANET) based on unmanned aerial vehicles (UAVs) are characterized by constant changed of the topology due to the moving of networking nodes. It causes an increased overhead costs for transmission of their coordinates when topology control is being done, or specific geographic routing protocols are being used. In order to reduce such overhead costs, a new approach based on medium access layer (MAC) adaptation is proposed. It evaluates the communication channel parameters, checks the integrity of the received data and manages retransmissions of information packets. Basic carrier sense multiple access/ collision avoidance (CSMA/CA) MAC layer protocol for the application of neuro-fuzzy adaptation have been considered and the types of FANET service messages to be optimized have been described. A flowchart of the new MAC layer neuro-fuzzy adaptation algorithm was shown. Simulation of FANET performance with neuro-fuzzy MAC layer adaptation, as well as classic CSMA/CA protocol, was provided. The simulation showed that the modified MAC protocol reduces the number of lost packets while having a comparable transmission rate and average latency. Thus, the simulation showed the effectiveness of the proposed method of neuro-fuzzy adaptation of the channel level of the UAV network