WSNs Frame Length Adaptive MAC Protocol Based On Q Learning

With the in-depth study of artificial intelligence and machine learning,the Q-learning algorithm has also been applied to wireless sensor networks(Wireless sensor networks,WSNs)medium access control(MAC)protocol,and has achieved certain results.The ALOHA-Q protocol combined with Q-learning is used in single-hop networks.Nodes set the frame length in advance and obtain dedicated transmission time slots for collision-free transmission through autonomous learning,which improves network performance.However,in the more common random spread multi-hop WSNs,the frame length cannot be set in advance,making the protocol unsuitable for the more general application scenarios described above.Aiming at the above problems,this paper proposes a Q learning based WSNs frame length adaptive MAC protocol.In this protocol,the network convergence is investigated when the relationship between the number of nodes and the frame length meets different requirements.The node sets the frame length according to the relationship between the number of one-hop neighbors,the frame length,and the network convergence.ALOHA-Q has the problem of pre-setting the frame length,and has verified the correctness of the frame length adaptive theoretical analysis through experiments,and by randomly selecting a node from its one-hop neighbors as the destination node in each frame,Reducing the probability of data packets accumulating at nodes with large frame lengths between nodes with communication links,solving the problem of data packet accumulation at the nodes found in the experiment,thereby optimizing the protocol and improving network performance.In this paper,NS-3 was used to simulate the proposed protocol and compared with ALOHA-Q and ALOHA protocol,which proves the applicability of this protocol in multi-hop networks and its performance advantages in terms of energy consumption and throughput.