Resource Modelling and Optimisation for SLA Guarantees in Hyperconverged Communication Infrastructures

超融合基础设施和虚拟化是通信服务提供商(CSP)关注未来数据中心和物联网架构的中心。计算、存储和网络成为按需和按需动态配置的商品化资源，以满足客户的SLA。多租户和多资源控制权限增加了SLA管理的复杂性。该项目将研究第一个虚拟网络功能并转移到其他资源类型，开发具有成本效益的分析模型，并结合创新的多目标优化方法来动态管理和调整资源分配，以实现SLA保证和最大限度的利用率。为了实现本项目的目标，需要完成的四项合作任务是：任务1：对NFV网络进行具有成本效益和效率的分析建模任务2：模型驱动的网络资源分配和优化任务3：通过模拟实验验证和改进算法任务4：在BT的测试床上或使用行业基准调查建议的解决方案

Hyperconverged infrastructures and virtualisation are at Communications Service Providers (CSP) centre of attention for the architecture of Future Data Centres and IoT. Compute, storage and network become commoditised resources dynamically provisioned on demand and at scale to meet customer SLAs. Multi-tenancy and multiple resource controlling authorities increase the complexity of SLA management. Studying first Virtualised Network Functions and shifting to other resource types, this project will develop cost-effective analytical models coupled with innovative multiple-objective optimization methods to dynamically manage and tune resource allocation in order to achieve SLA guarantees and maximize utilization. New solutions will be integrated with BT testbeds to investigate their viability.To achieve the aims of this project, four cooperative tasks to be completed are: Task1: Cost-effective and efficient analytical modelling for NFV networksTask2: Model-driven network resource allocation and optimizationTask3: Algorithms validation and improvement through simulation experimentsTask4: Investigation of the proposed solutions in BT's testbeds or using industry benchmarks