Load balancing is a significant aspect of cloud computing which is essential for identical load sharing among resources like servers, network interfaces, hard drives (storage) and virtual machines (VMs) hosted on physical servers. In cloud computing, Deep  Learning (DL) techniques can be used to achieve QoS such as improve resource utilization and throughput; while reduce latency, response time and cost, balancing load across machines, thus, increasing the system reliability. DL results in effective and accurate decision making of intelligent resource allocation to the incoming requests, thereby, choosing the most suitable resource to complete them.  However, in previous researches on load balancing, there is limited application of DL approaches. In this paper, the significance of DL approaches have been analysed in the area of cloud computing.  A Framework for Workflow execution in cloud environment has been proposed and implemented, namely, Deep Learning- based Deadline-constrained, Dynamic VM Provisioning and Load Balancing (DLD-PLB). Optimal schedule for VMs has been generated using Deep Learning based technique. The Genome workflow tasks have been taken as input to the suggested framework. The results for makespan and cost has been computed for the proposed framework and has been compared with our earlier proposed framework for load balancing optimization - Hybrid approach based Deadline-constrained, Dynamic VM Provisioning and Load Balancing (HDD-PLB)” framework for Workflow execution. The earlier proposed approaches for load balancing were based on hybrid Predict-Earliest-Finish Time (PEFT) with ACO for underutilized VM optimization and hybrid PEFT-Bat approach for optimize the utilization of overflow VMs.

Cloud computing is a rapidly growing technology in this new era. Cloud is a platform where users get charged on the basis of the services and resources they have used. It enables its users to access the cloud resources from the remote locations i.e. from anywhere at any time. It needs only a working internet connection to access the cloud services. Cloud users have always been victim to the security issues and attacks which leads to the data loss. The data is not saved on the hard disk of the computer so it is highly prone to security risks. Identifying the attacks on cloud platform is a difficult task because everything on cloud is in virtual form. EDoS (Economic Denial of Sustainability) attack is a form of DDoS attacks; carried out for a long span of time and intended to put a financial burden and cause economical loss to the users of cloud. Such attacks do not exhaust the bandwidth of the user; their main aim is to put a huge financial loss or burden on the user. A technique named as SPART (Supervised Pattern Attack Recognition Technique) implemented to mitigate the EDoS attacks in cloud computing which consumes lesser energy as compared to the existing models. The experimental results have shown the less energy consumption in proposed model.

Cloud computing is an emerging technology which provides unlimited access of versatile resources to users. The multifaceted and dynamic aspects of cloud computing require efficient and optimized techniques for resource provisioning and load balancing. Cloud monitoring is required identifying overutilized and underutilized of physical machines which hosting Virtual Machines (VMs). Load balancing is necessary for efficient and effective utilization of resources. Most of the authors have taken the objective to reduce the makespan for executing requests on multiple VMs. In this paper, a thorough review on scheduling and load balancing techniques has been done and different techniques have been analyzed on the basis of SLA Violations, CPU utilization, energy consumption and cost parameters.