Cross layer fault-tolerant cognitive diversity scheme for social information based artificial intelligent networks

Cross layer fault-tolerant cognitive diversity scheme for social information based artificial intelligent networks

Creators Zhong Yingji

Creators Ge Qi-Wei

Creators Matsuno Hiroshi

In this paper a special scenario of the topology in the artificial intelligent networks is studied, and a new cross-layer fault-tolerant cognitive scheme based on the topology information is proposed to exploit both the channel diversity and spatial reusability. The proposed Cross Layer Faulttolerant Cognitive Diversity Scheme (CL-FCDS) resides between the MAC layer and network layer, and aims to improve the network throughput by coordinating the transmission power at each node, selecting channel on each wireless interface, and deciding the routes based on the traffic information measured and exchanged among the multi-hop neighbor nodes. The Cognitive Diversity Air Time Metric (CDATM) is also introduced to quantify the difference of various adjustment factors. The proposed scheme achieves efficient utilization of available channels by selecting feasible adjustment candidates with the smallest CDATM value and coordinating the affected nodes to realize the adjustment. NS2 simulation reveals that the proposed scheme allows substantial improvement on the throughput and the outage probability, along with fault-tolerant cognition.