What is Voice Monitoring?

In essence, voice monitoring involves the continuous surveillance of network activities to assess the clarity, reliability, and overall effectiveness of voice calls. This is achieved by scrutinizing various parameters such as call quality, latency, jitter, and packet loss. By closely monitoring these metrics, organizations can identify potential issues and proactively address them to maintain seamless communication.

The telecom industry experiences higher churn than any other field. Therefore, the primary objective of voice monitoring is to uphold the highest standards in voice communication, guaranteeing that users experience clear and uninterrupted calls. This is particularly vital for service providers and businesses relying on voice-centric applications. Voice monitoring not only aids in detecting and resolving network issues, but also plays a pivotal role in adhering to Service Level Agreements (SLAs) and Quality of Service (QoS) standards.

The role of voice monitoring is becoming even more crucial in the age of VoIP. Voice over IP (VoIP) technology delivers significant benefits compared to traditional telephone lines. Unfortunately, all the benefits of VoIP can be quickly undone by poor call quality. Providers must therefore find ways to adapt while maintaining their quality of service. Often, this comes down to ensuring dual visibility, closely monitoring both the control plane and the media plane. The history of telephony puts emphasis on signaling monitoring and the connection between parties in the call.

However, with media now being transferred through ever more complex IP networks, service providers must ensure they have the tools and metrics to monitor RTP streams as well.

How Does Voice Monitoring Work?

Passive voice monitoring and active voice monitoring are two distinct approaches employed in voice monitoring practices, each offering unique advantages and functionalities. The primary distinction between passive and active voice monitoring lies in their level of involvement in the communication process. Passive monitoring observes ongoing communications without interfering, while active monitoring actively engages by generating synthetic traffic for assessment. Passive monitoring is valuable for continuous, long-term analysis, offering a comprehensive view of network performance trends. Active monitoring, on the other hand, is focused on immediate performance evaluation and is instrumental in obtaining a baseline for end-to-end service quality. Organizations often use a combination of the two approaches to gain a holistic understanding of their voice communication network’s health and performance.

Passive Monitoring

In passive voice monitoring, the system observes and analyzes voice communications without actively participating in the transmission process. It acts as a silent spectator, capturing and assessing data on the network’s voice traffic. 

This method is non-intrusive, allowing for real-time analysis of call quality, latency, and other performance metrics without affecting the actual communication flow. Passive monitoring is often used for comprehensive, long-term performance analysis and trend identification.

These are some of the pros of the passive monitoring approach:

• Continuous analysis of full live traffic instead of artificial samples
• View on actual customer experience
• Consideration of actual user devices and call paths
• Midpoint measurements that pinpoint problem sources

However, there are also some cons to it:

• Only implicit information on service availability
• Uncontrolled “experiments“ – not always possible to reproduce issues
• An end-to-end view can only be approximated

Active Monitoring

In contrast, active voice monitoring involves the deliberate generation of synthetic traffic or test calls within the network to assess its performance. This approach actively simulates real-world scenarios to measure specific parameters, such as call quality and responsiveness. 

Active monitoring is particularly useful for an end-to-end view, testing specific network components, and ensuring that the network can meet predefined quality standards. 

The following are some of the pros of passive monitoring:

• An end-to-end view
• Controlled “experiments“
• Insights into service availability
• Standards-based evaluation, e.g., PESQ and POLQA

The downsides, however, are:

• Artificial, e.g., no insight into end-device behavior
• Intrusive and potentially service-disrupting
• Consumes resources in live network
• End-to-end view does not say where problems occur

Voice Monitoring Best Practices

Listed below are some best practices:

1. Employing passive monitoring for network segmentation: VoIP calls traverse various network elements, facilitating the exchange of data between the calling and receiving parties. The network path connecting these parties usually encompasses a variety of components. Any of these network elements has the potential to introduce problems that may compromise service quality, thereby affecting the end-user experience. To pinpoint the specific network elements responsible for service degradation, it becomes necessary to partition the network into smaller segments. This segmentation approach is essential for identifying and isolating issues that may adversely impact the overall quality of service.
2. Correlating data from both the media and control plane during VoIP sessions: This process aligns the local signaling measurements (xDRs) and RTP media streams of a call to generate comprehensive call data records (CDRs), providing an intricately detailed perspective of the user experience. Specifically, xDRs encapsulate a summary of the call-related SIP signaling at a designated monitoring point, while CDRs encompass all the information derived from probes strategically dispersed throughout the network. This correlation method enhances the granularity of data analysis, offering a more nuanced and comprehensive understanding of the intricacies involved in VoIP sessions. Streams with the worst quality along the path determine the media quality of a call. Probes can track differences between streams or between xDRs for further analysis if metrics differ between streams or between xDRs.
3. Automating root cause analysis: Service providers can streamline their operations and notably diminish the time required for root cause analysis by leveraging automation. Recognizable patterns manifest in the data when encountering typical network issues, causing reproducible impacts on SIP and RTP traffic. Through the constant vigilance of probes overseeing the network, automated systems can promptly recognize and categorize prevalent impairment patterns linked to specific issues. This approach empowers service providers to proactively address and resolve common problems, contributing to a more efficient and responsive operational environment.
4. Utilizing targeted voice monitoring to conserve resources: Examine RTP packets in real-time and discarding them once pertinent information has been extracted. In cases where additional data is needed, activate probes to capture and archive media stream packets. This focused strategy in data capture combines efficiency with thoroughness. Unnecessary storage of irrelevant data is avoided, ensuring resources are used judiciously. However, when a comprehensive investigation is warranted, service providers possess all the necessary evidence for a thorough examination.
5. Creating statistics for trunks, routes, and other aggregates: Particularly for diagnostic purposes, having information on individual calls is crucial; however, call data alone does not offer a comprehensive overview of service performance. To gain insights at a broader level and not just focus on specific instances, it is imperative to utilize tools capable of aggregating data to produce meaningful and actionable statistics and Key Performance Indicators (KPIs).
6. Using the right metrics and KPIs: Service providers must delve beyond mere averaging of call quality metrics. They require data with elevated temporal resolution, capturing user experience throughout the entire duration of a call. Voice monitoring solutions equipped with time-slicing technology are accessible, employing real-time analysis of the RTP stream to precisely identify impairments as they occur during the call.

Which Voice Monitoring Solutions Are Available?

Voice monitoring solutions vary from traditional tools like Wireshark and HOMER to high-capacity, carrier-grade alternatives. Many standard tools struggle to handle intricate scenarios, providing only superficial insights by concentrating on RTCP and neglecting the core RTP. Typically, these tools center their attention on the control plane, emphasizing SIP signaling and call connectivity in a conventional manner.

Carrier-grade solutions offer a comprehensive approach to voice monitoring, thoroughly exploring aspects such as call management and in-call quality. These solutions grant precise insights into any voice service adhering to industry standards, enabling voice service providers to maintain smooth operations through extensive analysis of SIP and RTP traffic. By incorporating dual visibility—encompassing both SIP signaling and RTP media control—they provide a holistic perspective on VoIP. The systems automatically identify issues like dropped and silent calls, facilitating efficient performance optimization.

How to Choose a Voice Monitoring Solution?

Faced with a multitude of monitoring solutions, how can communication service providers determine which tools to employ? The following are six queries to consider when exploring potential network monitoring and voice analytics tools:

1. Can both SIP and RTP data be viewed by the solution?
2. Are the insights provided by the solution truly actionable?
3. What is the method for correlating all measurements of an incoming call?
4. What are the options for slicing and dicing the data?
5. Does the solution support on-demand packet recording?
6. What KPIs and metrics does the solution provide regarding service quality?