No abstract

No abstract

The virtualization and automation of network functions will be key features of future high-speed railway networks, which have to provide dependable, safe, and secure services. The virtualization of railway network functions will enable functions such as train control, train integrity protection, shunting control, and trackside monitoring and maintenance to be virtualized and to be run on general-purpose hardware. Network function virtualization combined with edge computing can deliver dynamic, low-latency, and reliable services. The automation of railway operations can be achieved by embedding intelligence into the network to optimize the railway operation performance and to enhance the passenger experience. This paper presents an innovative railway network architecture that features distributed intelligence, function cloudification and virtualization, openness, and programmability. The focus is on time-tolerant and time-sensitive intelligent services designed to follow the principles of service-oriented architecture. The interaction between identified logical identities is illustrated by use cases. The paper provides some details of the design of the interface between distributed intelligent services and presents the results of an emulation of the interface performance.

The emerging multimedia communications are expected to enable simultaneous services with high throughput and low-latency requirements. Ensuring appropriate quality of service (QoS) for multimedia services is one of the challenges that network operators have to cope with in the Evolved Packet System (EPS). The mechanism which allows policy based authorization and usage control of QoS resources intended for multimedia traffic in EPS is called Policy and Charging Control (PCC). The paper investigates the capabilities for open access to PCC. The open access allows third party applications to access functions in the telecommunication networks through Application Programming Interfaces (APIs). The support of standardized APIs for PCC is evaluated. An approach to design new APIs for PCC is described. APIs implementation issues are addressed including interface-to-protocol mapping. The functional verification of application server for open access to PCC functions is formalized. The application server's functional behaviour is modelled and adaptive load control mechanism is proposed. The application server performance and utilization are evaluated.

No abstract