Abstract:
Today, information networks play an important role in supply chain management. Therefore, in this article, clustering-based routing protocols, which are one of the most important ways to reduce energy consumption in wireless sensor networks, are used to optimize the supply chain informational cloud network. Accordingly, first, a clustering protocol is presented using self-organizing map neural network, SOM. Second, we cluster the network nodes based on two criteria of neighborhood and energy level using K-means clustering pattern. Third, we survey the efficiency and inefficiency of the clusters to balance the energy properly among the clusters. Then, to increase the network lifetime and to maintain the network DEA method is used. Finally, the model is tested for the information network of oil supply chain.
Machine summary:
Therefore, in this article, clustering-based routing protocols, which are one of the most important ways to reduce energy consumption in wireless sensor networks, are used to optimize the supply chain informational cloud network.
Therefore, we have measured the between the map unit of j and input sample i and t is the neighborhood function at the time of t, which is defined as follows:following repetition umber for SOM neural network training and mapping: ( t ) 0 exp t T (6)EPOCH=1000 rounds (3) Epoch is frequency number of the learning algorithm performance.
Research Methodology The considered project is a practical research which will be accomplished using scientific findings on neural networks and data envelopment analysis model regarding clustering the supply chain informational nodes and evaluating the cluster-heads efficiency.
Third, using information clustering methods and k-means in supply chain sensor networks, the extracted data of the neural network have been clustered with each cluster including several variables representing the lifetime and energy level of the clusters.
Input and Output Graph (the way of energy inputs and outputs to the main cluster-head) Research variables To develop the proposed model oil supply chain informational network has been surveyed and clustered into 1000 random nodes.
We evaluated 33 neural networks clustered by SOM and K-means algorithm using CCR model in data envelopment analysis.
We evaluated 33 neural networks clustered by SOM and K-means algorithm using CCR model in data envelopment analysis.