Mobile Internet, Cloud Computing, and M2M are considered major technologies that will bring great changes to the telecom industry. The growth of mobile Internet is the fastest. According to Ovum, by 2012, the number of mobile Internet users will have exceeded 2 billion, and with the mature commercialization of 3G and future 4G networks, mobile data bandwidth will be greatly improved. As a result, operator networks will run under greater pressure.
Operators today are faced with the problem of increased traffic but declining mobile broadband revenue. Low-value P2P traffic occupies a great deal of bandwidth; so to enhance user experience, operators need to invest more in expanding their network capacity.
The inherent richness and openness of many Internet services presents a challenge to service operation. If operators assume the position of mere “pipe” providers, they risk lowering their position in the telecom value chain. They will fail to benefit from fast-growing Internet services because they are not directly involved in service operation. A new operation model led by iPhone is taking a high position in the value chain, as shown in Figure 1.
It is therefore important for operators to adjust their business operation model and network architecture to adapt to rapid growth of mobile Internet. They need to optimize their operation and management systems to have a better knowledge of traffic distribution, service growth, and user preference, and to provide dynamic differentiated user and service management mechanisms. In this way, operators can make good use of network resources, can reduce CAPEX, and maximize bandwidth revenue rather than becoming mere pipe providers.
To optimize the operation and management of data services, a mobile core network must provide the following:
■ User and service based control and charging;
■ Dynamic control and charging based on differentiated terminal location, time, and usage;
■ E2E QoS control mechanism;
■ User behavior and traffic distribution analysis.
As traditional mobile packet core networks cannot provide these flexible, optimized functions, adding service identification, statistical analysis, service control, and content-based charging functions is necessary.
Service inspection is the basis for optimized operation and management. Services such as HTTP and P2P can be differentiated through deep inspection and analysis of L3–L7 protocols. The Deep Packet Inspection (DPI) technique is often used to inspect different protocols and applications via message feature codes and the traffic feature model.
Statistical analysis involves mobile data service and user behavior analysis. It provides network/user/traffic distribution, Top N users list, and network ranking. All these help operators deeply understand their networks, services, and user preferences, and help them optimize mobile Internet operation and management.
Service control contains policy management and execution. Policy management is intended to provide flexible, dynamic, real-time policies based on differentiated users, service, time, location, and traffic information. Policy execution includes users/services admission control, bandwidth restriction, resource reservation, and QoS control. These help operators make good use of network resources and achieve healthy business growth.
Unlike conventional charging based on time spent online or download/upload, content-based charging is a flexible billing method in which different services are charged at different rates. Each service is differentiated, and independent online/offline charging is offered to ensure that every bit of traffic in the network generates revenue.
To address the problem of increased traffic but declining revenue in mobile data networks, ZTE has launched ZOOMs―a complete optimized operation and management system that incorporates green, convergent, intelligent, broadband concepts into its 3GPP Policy and Charging Control (PCC) based architecture.
ZOOMs consists of four parts: packet core network gateway (GW/DPI), policy management system (PCRF/SPR), user behavior analysis system (UBAS), and online charging system (OCS/OFCS). GW/DPI provides service identification and control; PCRF/SPR implements user and service policy management; UBAS is responsible for user and network statistical analysis; and OCS/OFCS completes content-based charging. This system architecture is shown in Figure 2.
GW/DPI has a built-in DPI/DFI engine for L3–L7 protocol inspection. The DPI-based PCC mechanism implements dynamic content-based charging and service control, provides E2E QoS control together with the wireless network, and reports user web access log for service statistical analysis.
PCRF/SPR enhances the 3GPP PCC application mechanism. With DPI-based PCRF and SPR functions, it provides policy management and control in order to optimize wireless broadband network operation and maintenance. Policies are executed by GW/DPI based on user service subscription information and real-time user call information (including location, time, and traffic). These policies involve service admission, QoS control, charging, and traffic control.
UBAS performs complete service usage analysis, user behavior analysis, network traffic analysis, user preference analysis, and hotspot area analysis based on user and service information gained from network elements such as GW/DPI, HSS/AAA and Firewall. It provides operators with an effective means of analyzing and optimizing their networks and exploring new service opportunities.
OCS/OFCS provides online/offline charging and user account management functions based on 3GPP specifications. It also integrates a self-help portal server for management of user accounts.
ZOOMs provides dynamic policy and charging control, and flexible user behavior analysis, to help operators maintain complete control over both the service and user ends aspects of their network operations and to improve revenue. In this way, operators can gradually change their role from network operators to service providers, and take initiatives in mobile Internet development.