مقاله انگلیسی رایگان در مورد کاهش تأخیر در ماشین های مجازی – 2019 IEEE

I- Introduction and Motivation

II- Background and Related Work

III- Proposed Approach: Chain Based Low Latency VNF Implementation (CALVIN)

IV- Performance Evaluation of Elementary and Basic Functions

V- Evaluation of Computation- Intensive Advanced VNFs

VI- CONCLUSION

References

Abstract

Software-defined networking (SDN) and network function virtualization (NFV) processed in multi-access edge computing (MEC) cloud systems have been proposed as critical paradigms for achieving the low latency requirements of the tactile Internet. While virtual network functions (VNFs) allow greater flexibility compared to hardware-based solutions, the VNF abstraction also introduces additional packet processing delays. In this paper, we investigate the practical feasibility of NFV with respect to the tactile Internet latency requirements. We develop, implement, and evaluate Chain-based Low latency VNF ImplemeNtation (CALVIN), a low-latency management framework for distributed Service Function Chains (SFCs). CALVIN classifies VNFs into elementary, basic, and advanced VNFs; moreover, CALVIN implements elementary and basic VNFs in the kernel space, while the advanced VNFs are implemented in the user space. Throughout, CALVIN employs a distributed mapping with one VNF per Virtual Machine (VM) in a MEC system. Furthermore, CALVIN avoids the metadata structure processing and batch processing of packets in the conventional Linux networking stack so as to achieve short per-packet latencies. Our rigorous measurements on off-the-shelf conventional networking and computing hardware demonstrate that CALVIN achieves round-trip times from a MEC ingress point via two elementary forwarding VNFs (one in kernel space and one in user space) and a MEC server to a MEC egress point on the order of 0.32 ms. Our measurements also indicate that MEC network coding and encryption are feasible for small 256 byte packets with an MEC latency budget of 0.35 ms; whereas, large 1400 byte packets can complete the network coding, but not the encryption within the 0.35 ms.

INTRODUCTION AND MOTIVATION

Low latency communication is the central requirement for enabling the tactile Internet for human-machine co working [1]–[6]. Both, humans and machines require latencies below one millisecond for a wide range of co working scenarios. For instance, for humans operating in a virtual world and for interactions with robots and other machines, the latencies for visual, audio, or tactile multi-sensoric feedback should be below 15 ms, 3 ms, or 1 ms, respectively [7]. Every machine based on control loops also requires low latencies in order to work efficiently or to operate in a stable manner [8], [9]. As a concrete example, consider a classical inverted pendulum whose controller is placed in the cloud. Closing the control loop through a communication network will likely introduce some delays and packet losses. Fig. 1 shows the influence of the delay between the angle sensor and pendulum actuator (motor) on the pendulum stability. For long delays (50 ms in Fig. 1), the system becomes unstable, and the pendulum will never reach stability in the inverted position. For shorter delays (40 ms), the system takes some time to achieve stability. This time delay could imply lack of quality of service, and may affect other systems if the pendulum is part of a more complex environment with interconnected systems, or multiple pendulums coexisting in the same physical space.