|
| References | Resource provisioning techniques used | Major contribution | Pros | Cons |
|
| Karamoozian et al. [84] | Learning automata (LA) to reward or penalize the VM | QoS-aware resource allocation for media services | Optimal resource selection based on response time and failure rate | SLA violations, system throughput, and energy are not considered to ensure QoS |
| Zhang et al. [93] | Energy-efficient joint resource management and allocation (ECM-RMA) policy | Reduced time-averaged energy consumption in a multi-user multi-task in MCC | Improved QoS performance | Effect of device mobility to ensure QoS was not considered |
| Singh and Chana [85] | Q-aware: QoS-based resource provisioning | Analyzed cloud workloads and clustered using workload patterns | Significant reduction in cost and time | Unable to characterize the mobile client requests |
| Hassan et al. [86] | Cost-effective provisioning scheme for the multimedia cloud environment | Resource allocation and management based on the Nash bargaining solution | Efficient system in terms of utilization and reduced migrations | Execution time of the incoming request to be considered |
| Dabbagh et al. [87] | Integrated energy-aware resource provisioning | A greedy heuristic approach for generating a near-optimal solution using a simulated annealing technique | Effective for intensive computing and streaming | Scalability issues |
| Mitra et al. [88] | Mobility management system (M2C2) | Probing mechanisms | Supports mobility efficiently | All QoS metrics to be considered |
| Zhang et al. [89] | The resource allocation model is based on an auction mechanism | Combinatorial auction mechanism with substitutable or complementary commodities | Individual rational and incentive compatible | Does not fit the cloudlet architecture |
| Sood and Sandhu [90] | Proactive resource provisioning model | Independent authority to predict the future required resources using an artificial neural network | Achieves mobile user details and precision | Independent authority’s processing time and the communication cost are not considered |
| Din et al. [91] | Energy-efficient green solution | Hierarchical resource management based on novel 5G system architecture | Energy-efficient communication related to cost and time | Balancing the uneven energy consumption and traffic distribution required |
| Li and Xu [92] | Workflow can be executed by either the mobile device locally or the cloud server via computation offloading | Energy-efficient resource allocation algorithm (EERA) | Improved data communication time | Effects of dynamic bandwidth to be incorporated for a robust and adaptive system |
| Guo et al. [94] | Energy-efficient mobile edge computing systems | Computation-efficient models with a negligible and non-negligible base station | Optimally allocate the communication and computation resources | Cost ineffective for a large-scale geographic area |
| Avgeris et al. [95] | Efficient allocation of resources for the offloaded tasks from the mobile devices to the edge | Optimal resource allocation framework | Robust task offloading solution | Errors in dynamically estimated positions and end-user device numbers must be reduced to a minimum |
|
