A double layer access network architecture that provides Internet and Intranet for smart building services. The Internet provides network services, and the Intranet provides information exchange for private devices. At the same time, it supports dual-rate design, so there is no need to replace the device when the user rate is upgraded. This architecture simplifies network device requirement and improves sustainability. This paper mainly discusses the construction of audio-visual services on the optical fiber access network and the improvement of the provision method in the community, and service requirements can be flexibly adjusted.

As the growth of development on both network architecture and cloud computing, the existing quality of service (QoS) measurement mechanism in VNF design is not sufficient for the cloud network architecture, especially the development of container technology. It is important for service provider to provide to customers a service level agreement (SLA) on cloud network service by provision of Container-as-a-Service (CaaS) and container network function (CNF). In this paper, we propose an enhanced vQOS (EvQOS) measurement mechanism for container-based network architecture that retains the vQOS in VNF architecture [1] including the TWAMP light (TWL) protocol and the northbound interface design, and provides a containerization architecture and the friendliness of system management and monitoring.

The 5G era is officially coming. It bringing many new applications, such as self-driving cars, 3D VR, etc. Among them, edge computing technology is particularly critical which can reduce latency and increase throughput, thereby promoting the resources expansion and services diversity. This paper proposes to use high-speed network transmission technology RDMA and multi-access network transmission architecture to apply to multi-access edge computing including collaborative computing, data synchronization and data backup, etc. It is to provide low-latency and high-throughput network transmission so that edge computing users have better user experience.

In recent years, IP surveillance networks are expected to enable various practical applications, such as finding suspects, monitoring pedestrians, and securing societies (e.g., securing a city, a company and a data center). With these applications, IP surveillance network is regarded as one of the potential technologies toward developing smart cities. To support the concept of IP surveillance networks, automatic attribute recognition systems have emerged as a promising intelligent management system. To automatically recognize attributes of pedestrians (e.g., gender and clothing), we apply deep convolutional neural networks (CNNs), and the main contributions of this paper are threefold: (1) we proposed a practical system architecture for intelligent management of surveillance networks; (2) we implemented different deep CNNs, and an ensemble-learning method that leverages these multiple deep-learning models; (3) we evaluated the models using the real data of IP surveillance networks.

In recent years, due to customers have higher requirements for 4K/BK video and high-speed internet, telecom operators have begun to deploy FTTH network, but found that it is generally difficult to deploy fiber to the home, so G.fast technology has been favored by most telecom operators around the world and have begun to actively deploy. For the most widely deploy VDSL2 line with maximum rate that can only provide 100M internet service, a intelligent and accurate G.fast 300M high speed service prequalification technology, is a major research topic for telecom operators to promote 300M high-speed internet service. This paper proposes to use AI machine learning to estimate the G.fast line rate by using VDSL2 line attenuation to meet the real-site provision needs of telecommunications operators.

As the growing development of network architecture and Information Technology (IT), existing Quality of Service (QoS) measurement mechanism is not appropriate for the virtualization network architecture. It is important for service providers to satisfy the Service Level Agreement (SLA) of their customers when the network infrastructure is transiting to the next-generation. The popular method to verify the QoS parameters is using active measurement such as ICMP ping function and UDP Echo Plus message defined in [1], but both of them are lack of one-way latency measurement capability because they are not needed to time synchronizing on the two endpoints at first. In this paper, we will investigate the measurement mechanism and Network Function Virtualization (NFV) of home gateway nowadays, including components, protocol, and considerations. Then, we illustrate the procedure of Two-Way Active Measurement Protocol (TWAMP) defined in RFC 5357 [2] and development in the standard and market. Finally, depending on the requirements, such as network quality analysis and obstacle warning from service providers, we proposed our mechanism for the service providers to clarify the quality of service of subscribed lines of broadband access network, and decrease the complaint and increase the satisfaction of their customers.

The FTTx+VDSL2 access network architecture is one of the best choices for the broadband access network, and migrating to the FTTH access network for carriers. However, it may be difficult to build the fiber into/in the house for some situation, such like the ful-filled line pipe and the decoration issues. The carriers have to introduce next-generation copper-wire technologies for higher rate to form a hybrid access network in the FTTH bottleneck areas. In this paper, we had studied the various technologies that could provide high speed access network and enlarge the coverage of the broadband service, including G.fast, Vplus and XG.fast. Furthermore, we also proposed the test technologies and results to verify the G.fast technology.