October 11, 2005
Wireless LAN management - part 2
VoIP performance management is being pursued in the IP industry which is using a standards-based framework accepted by the IETF, ITU, and the ETSI. A distributed software-probe architecture is employed to provide call-quality feedback in real-time and in a cost-effective manner. The network is monitored for performance in real-time and detailed diagnostic data is provided.
The various network elements have QoS reporting protocols incorporated into them. The protocols are used for transmitting performance related information to the call control systems with only a minor increase in the traffic overhead. The call quality metrics can be assessed by the protocol RTCP XR. Call monitoring by using this framework enables quick isolation of faulty data so that it can be resolved in isolation before a transient issue becomes chronic. This is important because VoIP has a lower tolerance of delay and jitter than other IP-based applications.
Call-quality monitoring agents such as VQmon can be embedded in devices such as WLAN access points, switches, client devices, etc as the VoIP packet traverses the network path via these devices. VQmon agents offer the advantage of scalability and of being small in size and not making heavy demands of memory and CPU resources. Information regarding the listening quality scores and the severity of jitter along with its impact is provided by VQmon. VQmon executed these functions by using as less as 0.1% of the network bandwidth. This is because a very low reporting frequency.
The ITU G.107 E-Model with ETSI TS 101329-5 Annex E extensions forms the basis for VQmon, which can be used for the active testing of live calls. VQmon/EP (End Point) and VQmon/SA (Stream Analysis) are the two variants of VQmon. VQmon measures and reports the following parameters:
• Percentages of packets lost and packets discarded are reported. This helps to understand the extent to which a call is affected by network packet loss and jitter.
• By measuring the density of bursts, its impact on call quality can be understood and corrective actions undertaken.
• The length of time between bursts and the rate of packet loss during this period helps to understand user feedback in a better manner.
• The sources of prolonged delay can be identified by measuring the round-trip delay and end-system delay. This helps to devise plans in order to combat echo.
• The metrics of signal and noise levels help in pinpointing issues that originate due to a disproportionate disparity in signal and noise levels. This enables the detection of signal- and noise-level problems at mid-stream without having to decode the voice packets.
• Data obtained using the MOS and R factor can provide quick information on call quality. The R factor provides objective feedback on latency, jitter, packet loss etc for a given CODEC type.
• Incorrectly configured systems can be detected by using mid-stream probes.
Given below are the details of protocols supported by VQmon.
• RTCP XR is used for Media Path QoS Reporting and helps by enabling the collection and generation of call-quality reports by remote endpoints. Performance reports can pass through the firewall routers. It also helps to obtain analog signal information by means of network probes.
• H.460.9 Annex B, H.248.30 and SIP for Signaling QoS Reporting are used for call-quality reporting to call management systems. This helps the call-detail record (CDR) databases to provide information on the service quality to relevant customers.
• The SNMP and the RTCP XR Management Information Base (MIB) is being developed by the IETF. It will be used in the gateways and help in the recovery of metrics through SNMP. Call-quality information for both outbound and inbound streams will be available at the gateway.
--
Did you enjoy this post?
« Wireless LAN management - part 1 | Main | VoIP in Canada »
