Questions about Softswitch Network Design

1. Introduction

In recent years, Softswitch technology has been widely valued by the industry, four international standards organizations, ITU-T, ETSI, 3GPP, and IETF, have conducted research and standards development on Softswitch technology in different areas. The Standardization Organization of China has also done a lot of work. In general, standardization is at the forefront of the world. Several major carriers in China have also joined the construction of Softswitch networks. From the perspective of introduction, there are only two methods: one is to introduce long distance, optimize and divert existing VoIP services, such as China Mobile and China Weitong. One of the main representatives of the two traditional fixed operators is to introduce locally, quickly provide services, and seize the market. Due to the split between North and South, their respective game in the other's territory, Softswitch exactly provides technical means.

Deploying Softswitch throughout the entire network is not mature and feasible in the near future. This is not just a technical issue, it involves the current operating system, control and other aspects. This article mainly discusses several problems of softswitch networking, focuses on the analysis of softswitch network organization, IP bearer network organization, security design and number, and provides feasible solutions in the near future.

2. Softswitch Network Organization

The softswitch network includes two aspects: the routing design between SoftSwitch and the road network design. Routing design mainly refers to the signaling organization between Softswitch, generally using SIP-I protocol; the road network design mainly refers to the words of various media gateways road intercommunication.

2.1 Design of routes between Softswitch

Define the region of a vswitch as a control domain. In the routing design, the division of control domains is required. No matter what type of control domain division is adopted, multiple Softswitch involves addressing gateway devices, that is, routing organization. There are three routing organization modes: Full plane mode, hierarchical mode, and server locating mode. The full plane mode is used for the small size of softswitch at the initial stage. The classification method follows the mature multi-level Routing System of PSTN, which makes the routing data of each soft switch relatively simple and makes the structure of the softswitch network clearer. The locating server method has changed the previous two methods of signaling organization jump-by-jump transmission, and uses the IP accessibility to directly locate the peer soft switch. The signaling transfer is simple and clear.

Due to the poor interconnectivity between devices of current Softswitch manufacturers, the above three routing organization methods are not feasible in the near future, especially the server locating method is still in development. In actual networking, try not to introduce Softswitch throughout the entire network. You can introduce Softswitch from the local network or from the long distance, depending on the carrier's original intention of introducing SoftSwitch. In the signaling networking mode, the softswitch and traditional TDM hybrid networking methods are still taken into account. After the interconnection problem between manufacturers is solved, the softswitch classification or full-plane routing mode can be gradually introduced.

2.2 Traffic Network Design

Due to the adoption of the Fast IP forwarding technology, the soft-switch route organization is flat and connectionless. QoS and security should be considered in the design of the traffic network. Currently, the IP bearer network is mainly implemented using mpls vpn technology. To achieve communication between TG (relay gateway), AG (access Gateway), IAD (integrated access device), and smart terminals, a full network connection is required. The LSP (marking the switching path) of the MPLS logical link is a great challenge to the bearer network. Although VPN route reflectors can be used in the industry, however, the implementation of large-scale mpls vpn in the IP bearer network and the implementation of mpls vpn in the autonomous system (AS) still lacks support in practice. Recently, we recommend that you use a hybrid network with TDM to restrict the softswitch domain to a local or long distance. During local introduction, it is unrealistic to require the resident network to support the mpls vpn function because IAD and smart terminals are deployed on a large scale on the resident network, we recommend that you deploy service access control devices (SAC) at POP or settlement points in the Metropolitan Area Network to quickly converge IAD and smart terminal traffic and filter and forward real-time traffic.

3. IP Bearer Network Organization

3.1 Technical Requirements for IP bearer Networks

The Softswitch bearer network can be divided into three layers: the national backbone network, the Metropolitan Area Network and the resident network. Softswitch requires light load, security, and QoS. Because the real-time session-based services carried by the softswitch network are different from traditional Internet services, it is recommended that a national IP backbone private network be built separately to support the mpls vpn function. The man is complicated. According to the current operator's idea, the man is integrated, which requires the man to be able to distribute services. Man technology can be divided into four types:

* Large Metro IP network: the core and convergence layer of the metro must support the P/PE function. mpls vpn can be built directly. AG and SAC devices in the metro are directly connected to the PE router, you can also access PE devices through transmission networks or L2 VLAN links. SS (soft switch), SG (signaling gateway), and TG devices Implement mpls vpn intercommunication across the AS domain.

* Small Metro IP Network: in principle, local devices must support mpls vpn. If not, at least the edge layer can support the distributed PE function.

* Layer-2 Ethernet man: L2 switches connected to ag and sac must support VLAN isolation. Different users and business flows are isolated by VLAN; the man and PE devices in the backbone network are interconnected through VLAN trunk to associate VRF (VPN route forwarding table) from different VLANs to different VPNs, and access to different VPNs is achieved.

* MSTP man: AG and SAC connect directly to the PE device of the backbone network through MSTP, and connect the PE device to different vpn vrf through different MSTP access ports to achieve VPN access.

For resident network users connected by IAD, the IAD device must be able to support different marking of softswitch service traffic and Internet service traffic; SAC devices in the city must be able to identify the tags of IAD devices in the controlled area, distribute services, and import the softswitch and Internet business flows into different networks.

3.2 QoS of the IP Bearer Network

QoS of IP bearer networks is a hot topic in the industry. From the perspective of current technology, there are not many QoS methods to use IP networks to carry real-time Softswitch services. From the perspective of implementation, that is, the mpls exp field is used to ensure QoS of service priority. In terms of QoS of the IP bearer network, IPTN proposes adding a layer of bearer control layer to implement the establishment of mpls lsp. This method faces new problems. From the perspective of current technology, it is impossible to establish a dynamic LSP. mpls lsp channels can be established through static methods, in the future, the softswitch traffic dredging requires the mpls lsp to be fully networked, which severely restricts the implementation of MPLS Traffic Engineering. Further research and discussion are needed on how to achieve better QoS in the IP network.

3.3 Topology Design of IP Bearer Network

In order to adapt to the softswitch service load, the Topology Design Requirements of the IP network are more reliable. The following points should be taken into account in actual design:

* For link redundancy, a single node uses dual-link connections, with one master node and one slave node being set to an unequal value. When the main link is broken, you can use the link switching function to quickly switch to the backup link.

* In the node, consider the dual-route Device Design and load balancing. When a device goes down, you can quickly switch to another device.

* The core layer of the backbone network is designed with a dual graphic design approach and a full mesh connection for the link. The edge layer still adopts the dual attribution mode.

Related Articles]

• Softswitch: Future Analysis of profitable services
  

• Notes for Softswitch networking
  

• Next Generation Network (NGN): Research Progress of Softswitch Technology