The PTN bearer network has begun to be commercialized on a large scale in China Mobile Group, and the PTN pilot projects of China Telecom and China Unicom Group are also in full swing. In the actual operation and maintenance of PTN networks, how to change the traditional IP bearer network’s “best effort” characteristics or rough congestion management mode (such as simple FIFO queue scheduling, congestion tail drop, etc.), as well as formulate strategies for specific business QOS solutions for the current network, have become the focus of attention for operators at this stage.
While inheriting some essence of the traditional IP bearer network QOS technology, the PTN network has also developed a QOS strategy suitable for its own technical architecture. In view of the current network IP full service situation and network deployment strategy, the PTN network effectively uses various mechanisms to flexibly realize the refined management of the whole service, achieve carrier level end-to-end quality assurance, ensure users’ good network use experience and perceived recognition, and improve brand value competitiveness.
1. Several service models of QOS
QoS is an ATM performance parameter term that describes the transmission quality through a given virtual connection. These parameters include: CTD, CDV, CER, CLR, CMR and SECBR, ALL service classes, QoS classes, traffic contract, traffic control Quality of Service (QoS) refers to the quality of service. For network services, service quality includes transmission bandwidth, transmission delay, data packet loss rate, etc. In the network, measures such as ensuring transmission bandwidth, reducing transmission delay, reducing data loss rate, and delay jitter can be taken to improve service quality. Network resources are always limited, and as long as there is a situation of seizing network resources, there will be requirements for service quality. Service quality is relative to network services, and while ensuring the service quality of a certain type of business, it may be damaging the service quality of other businesses.
When it comes to PTN QOS, we have to mention several technical systems of Ethernet QOS, including:
Intserv model
IntServ (Integrated Services) initially attempted to categorize the services provided by networks into different categories on the Internet. The comprehensive service proposed by IEEE, IntServ, provides quality of service assurance for individual application sessions. IntServ defines three different levels of service types: (1) guaranteed service: providing a stable (mathematically provable boundary) for the delay of end-to-end packet queuing, making it possible to provide guaranteed delay and bandwidth services. (2) controlled load services. (3) best effort service: not providing any type of service guarantee.
Intserv can provide quality of service assurance for individual application sessions, which can be understood as reserving a certain bandwidth for a certain business traffic in the entire network and providing an end-to-end transparent channel for that business. This service type can provide complete QOS guarantee for business applications. However, this retention strategy will consume the already limited wide area network bandwidth, and when business traffic is not very frequent, it will cause bandwidth waste. Therefore, the scale of applications has been greatly reduced now.
DiffServ model
DiffServ is a network technology that ensures QoS. DiffServ defines an architecture that can implement scalable service classification on the Internet. A “service” is defined by certain important features when transmitting data packets through one or several paths in the same transmission direction within a network. These features may include quantified or statistical values of throughput, latency, latency jitter, and/or packet loss rate, as well as their relative priority in accessing network resources. The service classification is required to adapt to the needs of different applications and users, and allow the classification charging of Internet services.
This mode can be said to be the most widely used mode in the industry at present. IS210BPPCH1AD-IS200AEPAH1BHC, of course, also originated from the Intserv mode. It is a QOS strategy based on each hop. Each data device on the network classifies traffic by checking the header information of each packet, and then decides how to forward it based on the scheduling strategy. Therefore, network operators can set different QOS strategies on each data device according to the actual situation. Compared to the IntServ mode, the application level of DiffServ is simpler, easier to implement, and does not affect routing, resulting in higher efficiency. In terms of specific implementation, it includes functions such as traffic classification, congestion management, congestion avoidance, traffic supervision, and traffic shaping.
HQOS model
This is currently the most advanced QOS technology in the industry, and it can be said that it is also an improvement of DiffServ mode from single dimensional management to multi-dimensional scheduling. FiberHome Communication is one of the few manufacturers in the industry that can develop this QOS mode on PTN devices. In specific applications, this method can achieve differentiated scheduling and services for multiple users and businesses, and can also control the traffic of scheduling services simultaneously and schedule them according to the priority of the business. Compared to traditional QOS, some key parameters of QOS (such as CIR/PIR/WRED, etc.) are more flexible and convenient to set.
2. FiberHome Communication Full Service HQOS Solution
The emergence of PTN technology system is mainly to solve the problems of base station backhaul and IPization of large customer dedicated lines. IS210BPPCH1AD-IS200AEPAH1BHC reduces the cost of network unit bit and achieves end-to-end quality assurance with differentiated services. From the current situation of the current network, the services operated by the network are nothing more than VOIP (Voice over IP), HSDPA (Data Services), CES (Circuit Simulation TDM Voice), VOD (Video on Demand), IPTV (IP Television) and other types shown in the following figure.
For services with high latency requirements (i.e. the highest user perception), such as CES (TDM voice), VOIP, etc., EF (Enhanced Forwarding) strategy is adopted. When configuring CIR (Promised Bandwidth) and PIR (Peak Bandwidth), they are usually set to be equal (eg, CIR=PIR=5M). PQ (Strict Priority) or higher WRQ (Weight Balance) are used for queue scheduling to ensure service quality;