1. |
- Li, Jun, 1988, et al.
(author)
-
Bandwidth Slicing to Boost Federated Learning over Passive Optical Networks
- 2020
-
In: IEEE Communications Letters. - : Institute of Electrical and Electronics Engineers Inc.. - 1089-7798 .- 1558-2558. ; 24:7, s. 1492-1495
-
Journal article (peer-reviewed)abstract
- During federated learning (FL) process, each client needs to periodically upload local model parameters and download global model parameters to/from the central server, thus requires efficient communications. Meanwhile, passive optical network (PON) is promising to support fog computing where FL tasks can be executed and the traffic generated by FL needs to be transmitted together with other types of traffic for broadband access. In this letter, a bandwidth slicing algorithm in PONs is introduced for efficient FL, in which bandwidth is reserved for the involved ONUs collaboratively and mapped into each polling cycle. Results reveal that the proposed bandwidth slicing significantly improves training efficiency while achieving good learning accuracy for the FL task running over the PON.
|
|
2. |
- Li, Jun, et al.
(author)
-
Delay-aware bandwidth slicing for service migration in mobile backhaul networks
- 2019
-
In: Journal of Optical Communications and Networking. - : Institute of Electrical and Electronics Engineers Inc.. - 1943-0620 .- 1943-0639. ; 11:4, s. B1-B9
-
Journal article (peer-reviewed)abstract
- Fog computing is expected to be integrated with communication infrastructure, giving rise to the concept of fog-enhanced radio access networks (FeRANs) to support various mission-critical applications. Such architecture brings computation capabilities closer to end users, thereby reducing the communication latency to access services. In the context of FeRAN, service migration is needed to tackle limited resources in a single fog node and to provide continuous service for mobile end users. To support service migration, high capacity and low latency are required in mobile backhaul networks. Passive optical networks can be a promising solution for such mobile back-haul, in which bandwidth is shared by both migration traffic and that which is not associated with service migration. In this paper, we propose a bandwidth slicing mechanism, in which the bandwidth can be provisioned to the migration traffic and non-migration traffic dynamically and effectively to meet their different delay requirements. Simulation results verify that the proposed delay-aware bandwidth slicing scheme can handle the migration traffic properly, i.e., sending it within a required time threshold, while limiting the impact of the migration traffic on the latency and jitter of the non-migration traffic, particularly that with high priority.
|
|
3. |
- Li, Jun, et al.
(author)
-
Online Resource Management in Fog-enhanced Cellular Networks for Real-Time Vehicular Services
-
Other publication (other academic/artistic)abstract
- Fog computing is expected to be integrated with existing communication infrastructures, giving rise to the concept of fog-enhanced cellular networks (FeCNs) to support real-time services. In such FeCNs, service migration is necessary to maintain the service continuity and satisfy stringent latency requirements of real-time vehicular services, where the service is migrated from a source fog node to a target fog node following the vehicle's moving trace. Fog servers, however, need to have sufficient computational resources available to support such a migration. Also, provisioning resource for the migrated real-time services needs to be completed as soon as possible to minimize the service interruption. This paper proposes a distributed online resource management (ORM) scheme, in which resources for real-time vehicular services are provisioned with high priority. Once resources are scarce in one fog node, services with low priority can be migrated to neighboring fog nodes and their resources can be released in a distributed fashion. We propose two algorithms tailored to reduce the negative effects on the affected services. As a case study, the Luxembourg traffic volume model has been considered to verify the performance of the proposed scheme. Simulation results show that the performance of the proposed scheme is dependent on the backhaul capacity. Compared with other schemes, the one-hop access probability for real-time vehicular services implying low delay performance can be effectively improved, while the performance of other services can also be well maintained by providing sufficient backhaul capacity.
|
|
4. |
- Li, Jun, et al.
(author)
-
Service Migration in Fog Computing Enabled Cellular Networks to Support Real-Time Vehicular Communications
- 2019
-
In: IEEE Access. - : Institute of Electrical and Electronics Engineers (IEEE). - 2169-3536. ; 7, s. 13704-13714
-
Journal article (peer-reviewed)abstract
- Driven by the increasing number of connected vehicles and related services, powerful communication and computation capabilities are needed for vehicular communications, especially for real-time and safety-related applications. A cellular network consists of radio access technologies, including the current long-term evolution (LTE), the LTE advanced, and the forthcoming 5th generation mobile communication systems. It covers large areas and has the ability to provide high data rate and low latency communication services to mobile users. It is considered the most promising access technology to support real-time vehicular communications. Meanwhile, fog is an emerging architecture for computing, storage, and networking, in which fog nodes can be deployed at base stations to deliver cloud services close to vehicular users. In fog computing-enabled cellular networks, mobility is one of the most critical challenges for vehicular communications to maintain the service continuity and to satisfy the stringent service requirements, especially when the computing and storage resources are limited at the fog nodes. Service migration, relocating services from one fog server to another in a dynamic manner, has been proposed as an effective solution to the mobility problem. To support service migration, both computation and communication techniques need to be considered. Given the importance of protocol design to support the mobility of the vehicles and maintain high network performance, in this paper, we investigate the service migration in the fog computing-enabled cellular networks. We propose a quality-of-service aware scheme based on the existing handover procedures to support the real-time vehicular services. A case study based on a realistic vehicle mobility pattern for Luxembourg scenario is carried out, where the proposed scheme, as well as the benchmarks, are compared by analyzing latency and reliability as well as migration cost.
|
|
5. |
- Shen, Xiaoman, et al.
(author)
-
Heterogeneous LTE/DSRC Approach to Support Real-time Vehicular Communications
- 2018
-
In: 2018 10TH INTERNATIONAL CONFERENCE ON ADVANCED INFOCOMM TECHNOLOGY (ICAIT). - : IEEE. ; , s. 122-127
-
Conference paper (peer-reviewed)abstract
- Vehicle-to-everything (V2X) communications have become a key enabler for future autonomous driving by providing sensing and cooperation abilities beyond traditional on-board sensors. The main candidate technologies for V2X communications are dedicated short range communication (DSRC)/IEEE 802.11p and long-term evolution (LTE), each of which has its own strengths and weaknesses. DSRC provides low-latency direct connectivity, but suffers from serious link degradation and reliability issues. On the other hand, LTE is able to provide high capacity data transmission, while its latency remains an issue due to the long transmission time interval. Integrating these two radio technologies as a heterogeneous solution is promising to complement each other and support V2X applications with differentiated service requirements. This paper proposes a service-aware radio access technology (RAT) selection algorithm that enables a heterogeneous LTE/DSRC solution, where LTE and/or DSRC are selected according to services. Each vehicle is assumed to be equipped with both LTE and DSRC interfaces. The heterogeneous solution selects RAT based on the services requirements with consideration of the networks performance. The solution is implemented with integration of the C-ITS standard and validated through the OMNet++ and SUMO based network simulator. A see-through scenario, where a fleet of vehicles require real-time video data sharing to assist efficient driving, is considered. The simulation results demonstrate that the proposed approach outperforms the purely DSRC approach significantly and is able to simultaneously support safety message transmission with low latency, video streaming with high bandwidth requirements, and other applications with different service requirements.
|
|
6. |
- Shen, Xiaoman, et al.
(author)
-
Minimizing Registration Overhead for Multipoint-to-Multipoint Communication in Passive Optical Interconnects
- 2017
-
In: 2017 OPTO-ELECTRONICS AND COMMUNICATIONS CONFERENCE (OECC) AND PHOTONICS GLOBAL CONFERENCE (PGC). - : IEEE. - 9781509062935
-
Conference paper (peer-reviewed)abstract
- To facilitate the conflict-free multipoint-to-multipoint communications in data center networks, we propose and evaluate two registration schemes, which are able to utilize the minimum amount of control frames exchange for the polling based medium access control protocol in passive optical interconnects electronic.
|
|