Research On Anti-jamming MAC Technology In Wireless Ad Hoc Network

A wireless Ad Hoc network is a dynamic network which can be established without relying upon communication infrastructures.This kind of network has the characteristics of self-organizing,centerless,and multi-hop etc.However,it is vulnerable to various jamming techniques.Usually,the purpose of a jammer is to downgrade the quality of data transmission among nodes.Traditional anti-jamming strategies usually need a large overhead.Recently,the proposed intelligent jamming attacks can point at weaknesses like waveforms,protocols,etc.in communication systems to achive a better jamming effect with a lower overhead.Traditional Media Access Control(MAC)technology is designed to enable multiple nodes to efficiently use the shared channel.This technology only considers the internal interference between the nodes,but does not consider the jamming from the outside.For this purpose,this paper does a research on the channel access process of nodes,and putforward two anti-jamming schemes based on MAC layer pointing at different jamming strategies.The main work is as following.An anti-jamming algorithm based on Markov Decision Process(MDP)is proposed for the network anti-jamming problem with known jamming patterns.With this algorithm,the transmitting node obtains all the network states under the specific jamming pattern by learning in advance and gets the optimal channel selection strategy under the specific state.In the communication activity based time slots,the node selects the optimal policy corresponding to the current state according to the prior knowledge.In the process of selecting different policies according to different network states,channel hopping is involved,which will make a node cost more and reduce the network performance.Therefore,when modeling the channel selection process as MDP,this paper considers channel hopping overhead,encouraging the node to stay in channels with good quality to achieve efficient avoidance of jamming while improving the network performance as much as possible.Results show that in the network with the determinate jamming pattern,this algorithm performs better at the message delivery rate than the random channel hopping anti-jamming strategy and the fixed channel transmission strategy,and this algorithm still achives a good message delivery rate in the scenario of appropriately slowing the signal hopping speed.An anti-jamming algorithm based on Multi-Armd Bandits(MAB)is proposed to solve the problem of network anti-jamming under adaptive dynamic interference.With this algorithm,the node learns dynamically through "exploitation" and "exploration" in the process of resisting the jamming attacks,grasps the state of each channel and selects the best channel access strategy.It is not possible to maintain high quality for channels due to adaptive jamming conditions.With this algorithm,the node continues to explore channels that are not frequently accessed while exploitating the high-quality channel for communication,can adjust the channel selection in time when the original good channel is degraded.While defending against jamming attacks,the algorithm can also equalize the traffic carrying capacity of the channel and reduce the probability of multiple access jamming between transmitting nodes.This algorithm is simulated and analyzed.The results show that this algorithm can achieve a better message delivery rate than random channel hopping anti-jamming strategy and fixed channel transmission strategy when facing different jamming attacks.At the same time,multi-access interference between nodes under different anti-jamming strategies is simulated and analyzed.It is found that in the adaptive jamming environment,this algorithm achieves a lower probability of multiple access interference than the random channel hopping anti-jamming strategy.