Efficient admission control and resource allocation mechanisms for public safety communications over 5G network slice
Autor: | Anuar Othman, Nazrul Anuar Nayan |
---|---|
Rok vydání: | 2019 |
Předmět: |
Service (systems architecture)
Computer science Heuristic (computer science) business.industry 020302 automobile design & engineering 020206 networking & telecommunications 02 engineering and technology Admission control Task (project management) 0203 mechanical engineering Scalability 0202 electrical engineering electronic engineering information engineering Cellular network Resource allocation Reinforcement learning Electrical and Electronic Engineering business Computer network |
Zdroj: | Telecommunication Systems. 72:595-607 |
ISSN: | 1572-9451 1018-4864 |
DOI: | 10.1007/s11235-019-00600-9 |
Popis: | The deployment of a broadband public safety (PS) mobile network can be undertaken in different ways. One method involves combining commercial networks with private ones to reduce their deployment cost and time to market. In shared commercial networks, priorities should be defined to differentiate traffic not only between consumers and PS users but also between different PS organizations and type of services. Prioritization must also ensure that emergency calls are always served under normal conditions and during disasters. The recent advent of the fifth generation (5G) wireless standard introduces new technologies, such as network slicing (NS), which allows the provision of logical PS networks in a shared 5G system wherein each slice can be dedicated to an organization or to a type of service. However, 5G management and orchestration become a challenging task with NS, e.g., in handling resource allocation between slices with diverse requirements. Therefore, efficient solutions for slice resource allocation are required to facilitate this task. In this paper, we present a review of adaptive and dynamic resource allocation leveraging on heuristic and reinforcement learning-based algorithms that have been proposed in the recent literatures. The challenge in implementing these algorithms is to find the most suitable one for our problem, i.e., an algorithm that is highly scalable, able to solve problems immediately, and exhibits the best convergence properties in terms of speed and ability to find the global optimum. |
Databáze: | OpenAIRE |
Externí odkaz: |