This special issue contains a collection of invited and contributed papers on low latency in edge optical networks. We present a brief introduction followed by an overview of the topics covered in the different papers.
© 2019 Optical Society of America
Higher capacities as well as lower capital and operational costs have typically been the drivers of the evolution in the optical transport layer. Recently, latency requirements are exerting increased influence in optical network design. While low latency has always been a desirable attribute, it is now becoming a must-have for networks moving forward. This is particularly true for edge networks as a growing number of applications rely on low latency to be viable and computing is moving closer to the users.
Mobile fronthaul networks, for example, have very strict latency and timing requirements. Human-in-the-loop (HITL) applications, such as augmented reality or haptic communications, are driven by the low response time that is tolerable to humans. Multi-access edge computing (MEC) and fog computing encompass distributed computing at the network edge; low latencies among the edge entities are needed to be more efficient than remote-cloud-based implementations. In this special issue, two invited and seven contributed papers investigate different aspects of low latency in edge optical networks.
The first invited paper “Towards Immersive Tactile Internet Experiences: Low-Latency FiWi Enhanced Mobile Networks with Edge Intelligence” focuses on the emerging Tactile Internet as one of the most interesting low-latency applications. It describes how fiber-wireless (FiWi) access networks and edge computing can be used in teleoperations to provide quasi real-time haptic feedback to a human operator controlling a remote robot. The second invited paper “Transport Evolution for the RAN of the Future” describes how 5G imposes new requirements on the optical transport network. Starting with a description of use cases, it reviews current standardization activities, explains the radio access network (RAN) architecture, and discusses optical transport solutions to meet the capacity, latency, and timing challenges of 5G.
The contributed papers span a range of different topics. The paper “Delay-Aware Bandwidth Slicing for Service Migration in Mobile Backhaul Networks” discusses a delay-aware bandwidth assignment for passive optical network (PON) based mobile backhaul networks. The mechanism allows the segmentation of delay-insensitive large data transfers in order to minimize the impact on delay-sensitive mobile traffic. In the paper “Low-Latency and Energy-Efficient BBU Placement and VPON Formation in Virtualized Cloud-Fog RAN,” the authors present a two-tier edge cloud architecture using a time and wavelength division multiplexed passive optical network (TWDM-PON). Employing different strategies for virtual baseband unit (BBU) placement and PON connectivity, the trade-offs between propagation latency, power consumption, and blocking probability are investigated. The paper “Decoupling of Uplink User and HARQ Response Signals to Relax the Latency Requirement for Bridged Fronthaul Networks” advocates a separation of the transmission of uplink user data from hybrid automatic repeat request (HARQ) responses. Giving the uplink user data lower priority, the downlink latency can be improved in time-division duplex schemes (TDD). In the paper “Minimizing Delay and Packet Delay Variation in Switched 5G Transport Networks,” the authors present a novel scheme to asynchronously aggregate Ethernet-based fronthaul traffic. Combining a time window with a time-out approach, bounded delay transmission can be obtained. The paper “Latency-Aware CU Placement/Handover in Dynamic WDM Access-Aggregation Networks” provides an adaptive latency-aware algorithm for a dynamic placement of central units (CUs) in optical access-aggregation networks. The algorithm also performs grooming, routing and wavelength assignment of mobile network traffic demands. In the paper “Latency-Aware Resource Orchestration in SDN-Based Packet Over Optical Flexi-Grid Transport Networks,” the authors investigate the service chaining of virtualized network functions across distributed data centers taking latency, bandwidth, and computing requirements into account. Finally, an integrated network resource management scheme to best allocate cloud resources available at both remote radio head (RRH) and BBU locations is presented in the paper “Hierarchical Edge Cloud Enabling Network Slicing for 5G Optical Fronthaul.”
A big thank you goes to all the authors for submitting their work to this special issue and to all the reviewers for their valuable and timely feedback. Particular thanks go to JOCN Editor-in-Chief Jane M. Simmons. Without her energy and hands-on support, the realization of this key special issue for the evolution of edge optical networks would not have been possible.
Lead Guest EditorRamon Casellas
U. Carlos III of Madrid, Spain
Nokia Bell Labs, Germany