The Industrial Internet of Things (IIoT) can transform an existing isolated industrial system to a connected network. The IIoT and the related wireless connectivity requirements for industrial sensors are very significant. The deployed sensors in IIoT monitors the conditions of the industrial devices and machines. Therefore, reliability and security become the most important concerns in IIoT. This introduces many familiar and ever-increasing risks associated with the industrial system. The IIoT devices can be vulnerable to vast array of viruses, threats, and attacks. Therefore, an efficient protection strategy is required to ensure that the millions of IIoT devices are safe from these risks. However, resource constraint IIoT devices have not been designed to have effective security features. Due to this, in recent years, cloud, fog, and edge-based IIoT has received great attention in the research community. The computationally intensive tasks such as security, data analytics, decision making, and reporting are performed at the cloud or fog using a powerful computing infrastructure. The data security of the IIoT device has been provided by employing improved Rivest-Shamir-Adelman (RSA) and hash signatures. The proposed RSA algorithm has a four-prime number of 512-bits. The device authentication is performed by employing a hash signature. For long network life, an efficient clustering technique for the sensor devices which is based on node degree(N), distance from the cluster(D), residual energy(R), and fitness (NDRF) has been proposed. The fitness of the sensor nodes is computed using the Salp swarm algorithm (SSA). In order to reduce the latency and communication overhead for IIoT devices a resource scheduling using SoftMax deep neural network (DNN) is proposed. All the requests coming from the cluster head are classified using SoftMax-DNN for best resource scheduling on the basis of storage, computing, and bandwidth requirements. The proposed framework produces superior results, especially in terms of energy consumption, latency, and strength of security.