Two Methods for enterprise VoIP network call control

QoS can give priority to voice communication services under limited network resources. However, if too many voice communications occur, the network may be congested. How can this problem be solved?

Call Control for enterprise VoIP networks

When we configure the QoS function that supports voice on an enterprise's IP network router and switch, this IP network is ready to carry the voice application. However, in order to achieve the call, we also need to establish a correct voice call route, while also facing the problem of voice congestion: QoS can distinguish between voice traffic and other data traffic, however, voice traffic congestion cannot be controlled.

Call Control

Compared with carriers, the call routing design of Enterprise VoIP networks is relatively simple, especially in the WAN aspect. Currently, call control methods commonly used in enterprise VoIP networks are as follows: dial-peer and H.323.

The Dial-peer method is easy to use, but each Voice Gateway Router must be configured with a call route relative to other voice gateway routers. When the number of vrouters is small, it is very simple and the WAN resource overhead is minimal. However, when there are a large number of Voice Gateway routers or New Voice Gateway routers are added, You need to configure each Voice Gateway Router, which has poor scalability.

It adopts the control protocol of H.323 gateway and network guard to manage call routes, which has good scalability and simple network configuration. Each H.323 Voice Gateway Router registers its own call information to the network guard through the RAS signaling when accessing the network. In this way, all the call control information appears on the network guard. The Voice Gateway Router needs to send a request to the network guard when establishing a call connection to obtain the relevant information of the target Voice Gateway Router, thus establishing a call connection.

Both Dial-peer and H.323 call control are very mature technologies, and call permit control in VoIP networks is relatively complicated.

The problem of the Call permit Control (CAC-Call Admission Control) is caused by the Best-effort and non-connection-less attributes of the IP network. For a traditional connection-oriented telephone network, when a call passes through each telephone switch, if the switch finds no circuit to provide a call connection, the call is rejected, this does not affect established call connections. In a VoIP network, the Voice Gateway Router does not know the network conditions, including whether the network's Link resources are saturated and network latency, whether the packet loss rate and network jitter parameters can meet the needs of voice applications. If the network cannot provide sufficient resources to ensure the voice quality, creating a new call link may affect the established link or reduce the quality of all calls. If the bandwidth of the WAN can only accommodate two voice calls, the establishment of the third call will lead to a reduction in the quality of all calls. In this case, the call denied has established a link without quality assurance.

The network QoS mechanism we discussed in the previous article ensures that voice traffic is not disturbed by other network traffic. It can only control the existing traffic on the Internet. The call permit control ensures that the voice traffic in the network is not disturbed by other voice traffic and controls the traffic entering the network. Call permit control also allows the Voice Gateway Router that initiates a call to seek other call routes after confirming that the call fails. For example, you can use the public telephone network as a backup route to ensure that the telephone network is smooth.

Compared with connection-oriented circuit switching, in non-connection-oriented IP networks, it is not easy to determine whether the network has resources to provide voice services, the TCP/IP protocol standard does not provide a corresponding mechanism. Although there is no perfect solution yet, there are still several ways to effectively solve this problem. These methods can be divided into three types based on different mechanisms:

Call permit control based on the Local Voice Gateway status: The call permit control based on the Local Gateway status is based on the configuration information of the gateway, such as the maximum number of calls of the corresponding gateway; if the number of calls to the corresponding gateway exceeds the configured number, the caller gateway can reject the call. In addition, the gateway does not have network resources to other gateways when the network interface status changes. The Gateway can reject the call. This method is easy to use without increasing the overhead of network resources. However, when the network topology is complex and the network status changes, it may become invalid.

Call Permit Control Based on Network Status: Call permit control based on network status is a method used by the gateway to confirm that the network from the gateway to the called Gateway meets the voice communication requirements before initiating a new call. Currently, the most common method is to use the network status detector built in the Gateway Router, such as Cisco's SAA (Security Assurance Agents ), the detector gateway can determine the network delay, packet loss rate, jitter, and other parameters. The Gateway can determine whether to establish a call connection based on these parameters. This method is very effective and can adapt to various network scales and have good scalability, but it has high requirements on gateway devices.

Call Permit Control Based on Network Resources: Call Permit Control Based on Network Resources uses a mechanism to reserve network resources before a call is established to implement end-to-end network assurance. If it is determined that the network does not have enough resources, the gateway should refuse to establish a call connection. Currently, the most representative method is to reserve network resources through H.323 calls and RSVP signaling to ensure the quality of speech.

This method provides end-to-end service quality assurance and call control, but requires all network devices to support the RSVP protocol. This poses a challenge to the requirements of enterprise IP network devices and network expansion, this increases the signaling overhead of the network.

The three call permit control methods described above are not isolated. When an enterprise's VoIP network structure is complex, multiple methods may be combined to solve the call control.

When a vro in an enterprise's IP network is configured with the correct QoS parameters, it provides the voice call permission control function on the corresponding gateway and network guard, telephone services based on enterprise IP networks can completely replace traditional telephones.

1. Latest developments in the global enterprise VoIP Market
2. Implementation and network organization of Enterprise VoIP