Two-fold spatiotemporal regression modeling in wireless sensor networks

摘要: Distributed data and restricted limitations of sensor nodes make doing regression difficult in a wireless network. In conventional methods, gradient descent Nelder Mead simplex optimization techniques are basically employed to find the model incrementally over Hamiltonian path among nodes. Although based approaches work better than ones, compared Central approach, their accuracy should be improved even further. Also they all suffer from high latency as network traversed node by node. this paper, we propose two-fold distributed cluster-based approach for spatiotemporal networks. First, regressor each cluster is obtained where spatial temporal parts cluster's learned separately. Within cluster, collaborate compute part head then uses particle swarm learn part. Secondly, heads apply weighted combination rule distributively global model. The evaluation experimental results show proposed brings lower more energy efficiency its counterparts while prediction considerably acceptable comparison with approach.