Simulation and Analysis of Mathematical Methods in Real-Time Engineering Applications. Группа авторов
edge to overcome this challenge by taking user and base station as requirements. The DRL helps reduce power and bandwidth from the base station to the user, thus making the system energy efficient [33].
The main aim of the DRL method is to provide energy efficiency and a better user experience. Another advantage of DRL is that it has the capability not to exceed the space of the base station. Convex optimization method is first derived to obtain minimum transmission energy and iterate with DQN. It also reduces the space state of the network. On the basis of convex optimization results, optimal connection and optimal power distribution are found. Agent and external environment are the two states of DRL. By taking different actions, the external environment state is achieved. The external environment receives a reward. The main purpose remains as to maximize the value of the reward. In the experimental analysis, several users with three base stations are considered. The number of users for each convergence step is considered. It is seen that as the number of users increased, DRL required more steps for convergence; thus, convergence speed tends to slow down and the efficiency has also increased [33].
Table 2.1 Existing studies using deep learning in edge.
| S. no. | Existing methods | Inference |
|---|---|---|
| 1. | Joint task allocation and Resource allocation with multi-user Wi-Fi. | To minimize the energy consumption at the mobile terminal, a Q-learning algorithm is proposed. In this method, energy efficiency is not considered, which leads to additional costs for the system. |
| 2. | Joint task allocation-Decoupling bandwidth configuration and content source selection. | An algorithm was proposed for avoiding frequent information exchange, which was proven to be less versatile and hence cannot be used in large applications. |
| 3. | Fog computing method for mobile traffic growth and better user experience. | As users are located in different geographical places, implementing fog becomes challenging and requires high maintenance and increased costs. |
| 4. | Deterministic mission arrival scenario | After successfully completing the present mission, each mission is completed, which cannot work as the data source generates tasks continuously, which cannot be handled by the deterministic method. |
| 5. | Random task arrival model | This method works on task arrived and not on the queue tasks, which fails the system to work efficiently. |
2.5.3 Computation Offloading Using Deep Learning
Computation offloading is a great mechanism to offload extensive tasks at the nearby server and communicate cloud with important/filtered data. With edge, computation offloading has excellent applications for mobile devices by enhancing efficiency.
In a study, a dynamic computing offloading mechanism is performed. The objective of the study was to reduce the cost of computational resources. Mobile edge computing is considered (MEC). A Deep Learning method, i.e., Deep Supervised Learning (DSL) is considered. A network of a mobile-based computer system is considered. A pre-calculated offloading solution is proposed. A continuous offloading decision problem is formulated as a multi-label classification problem. After experimental analysis, it is inferred that as the exhaustive strategy suffers exponentially with the increase in the “n” fine-grained components.
2.6 Evolutionary Algorithm and Edge Computing
Usually, an evolutionary algorithm is used to solve the NP-hard problem, where solving the problem by a traditional optimization mechanism is impossible. This evolutionary algorithm takes some ransom solution vector from the solution space and tries to get optimal solution by the n number iteration by slowly evolving towards an optional one in each iteration through some cost or reward function. There are many evolutionary algorithms like 1) particle swarm optimization, 2) Genetic algorithm, 3) colony optimization algorithm.
In edge computing, many NP-hard optimization problems could be solved using those evolutionary algorithms.
In Mobile edge computing (MEC), offloading makes low latency and energy-efficient. Security critical tasks involve more computation and take more time. If we offload them, we can achieve a good performance. To minimize task completion time and energy consumption, particle swarm optimization algorithms are proposed [35]. Position-based mapping is carried out to map the particle solution of scheduling. a new slow down particle movement process mechanism is reported in the update particles step of the algorithm. Tthey have proved that the new update mechanism with slow down process achieved better performance compared to that of conventional particle swarm optimization algorithm.
Another offloading mechanism in a mobile edge environment is proposed with a combination of queuing network and genetic algorithm for the mobile edge environment [36]. Predicting the waiting time and service time of the edge server is an important one to make offloading decision. A queuing network model is introduced in this work to model the waiting time and the service time. The waiting times and the service times which are generated from the queuing network are taken as an indirect indicator for the load level of the edge server which is used for the offloading decision by using genetic algorithm. The genetic algorithm is designed to make optimal offloading by minimizing the task response time with the constrain of load level of the edge server and task transmission time from node to edge server. This proposed algorithm outperformed compared to that of particle swarm optimization and traditional round-robin scheduling based offloading in terms of response time.
Using the vehicular network node as an edge node is the latest trend. Figure 2.8 shows the vehicle edge network. In Figure 2.8, the roadside sensor node will offload the complex computations to the vehicle computing node. When the vehicle moves, it transfers the load to another vehicle to carry over the computation. Once the computation is finished, it will offload to the roadside sensor node.
Figure 2.8 Offloading in vehicular node.
There are a few research works done to utilize the vehicle as an edge node. An edge computing model for vehicle network applications to improve vehicles’ computational capacity is introduced with a scheduling algorithm using ant colony optimization [37]. This framework makes it possible to use a dynamic vehicle network and the vehicle as an edge service server and improve the connected vehicle application’s computational power. The autonomous organization of vehicle edges is also addressed in this work. The proposed job scheduling on the vehicle edge nodes provided excellent performance in urban and highway scenarios where many vehicles’ presence offers more optimization solutions.
2.7 Conclusion
The chapter has given a basic introduction to edge computing with its associated research challenges. The various mathematical models for solving the edge computing problem are also explored. An insight on computational offloading and multiple approaches for computational offloading are discussed. The Markov chain-based decision-making approach is an efficient mathematical approach. The applicability of it for the computational offloading problem in edge computing is also explored. A game theory-based dynamic approach for offload decision making is provided with available solutions. Achieving target