Download PDF
Monitoring IPTV For Service Providers
IPTV and Video over IP services are emerging from many providers all around the world. Service Providers are planning and deploying IPTV as part of the Triple Play, Voice, Video and Internet, service to the home. The competition is and will increasing become fierce as all Service Providers like cable, satellite and Telcos big and small, plan to protect and grow revenue from Triple Play Services. Video, Quality Video, is the key to that growth and longer-term survival of the Service Provider. Voice and Internet Services will not be enough to compete when products like High Definition, local and national sports packages, video on demand, etc, are offered by the competition as a bundled Triple Play product to the home.
Unlike voice and data there is almost zero tolerance for poor or “spotty” video service. Because consumer tolerance for poor quality video service is very low, the highly competitive Triple Play market, and magnitude of failure to deploy Triple play so high for the Service Provider, monitoring and tools for quality must be considered top priority.
There are five reasons why monitoring IPTV will be required for the success of IPTV deployments and Triple Play (Voice, Video and Internet) services to the home as a whole.
1. Knowledge of Outage
Knowing when a video service is out or delivering poor quality is critical to solving the problem. Whether a Service Provider knows there is an outage or not the customer will know. The customer may or may not call, either way reputation, quality and business will suffer. Video is 24/7 on hundreds of streams and the service providers need to know what is happening throughout the network on all streams at all times.
2. Magnitude of Outage
Monitoring tells the Service Provider how big of an issue has occurred or is occurring in the network. Large outages need quick attention as the customer will call very upset for each minute the service is down. Subtle outages, such as periodic video pixelization and/or audio artifacts, may not be enough to cause a call from the customer but goes to the underlining perception of quality. These issues can simply cause reductions in subscriptions again affecting the stability of the Triple Play business as a whole.
3. Scope of Outage
Monitoring instantly tells the Service Provider the scope or size of the outage. If the outage originated from the Headend every customer in deployment is affected, say 20,000 to 500,000 people. If the outage is off the ring network at the Central Office (CO) or Hub site then only the customers off the CO/Hub will see it, say 1,000 to 5,000. And if the problem happens at a DSLAM/QAM then the customers off the DSLAM/QAM will see it, say 100 to 500 people.
4. Prevent Outages
Monitoring has the ultimate goal to prevent the outage and deliver superior IPTV service 24/7. Knowing the limits and behavior of the IPTV network (Converged or not) helps the Service Provider understand system thresholds and patterns to monitor for. These thresholds can trigger to alert the Service Provider to system instability before there is an outage.
5. Cost of Not Monitoring
Customers will not continue to pay for IPTV if the service has too many outages and/or outages that take too long to fix. They will simply leave. The reputation of poor quality will also affect sales of future customers. With the increase of completion for video services, some other Service Provider will take that business. More than just IPTV, they will take the entire Triple Play revenue for that customer. The cost of poor Quality of Experience (QoE) is enormous, monitoring and tools for the deployments are critical to the effort of QoE.
How do we get there from here?
IPTV deployments and the technology are in their infancy. There are few standards and best practices are being experimented with in the field. A typical IPTV deployment consists of at least one Headend where video is gathered, groomed and injected into a Metro (and WAN) network that has drops in different geographic dispersed service areas at CO or Hub sites. The video service passes from the CO, down to the set top box at the home through fiber, to homes last mile networks, DSLAMs for DSL last mile networks or QAMs for RF last mile networks.
Given this topology and state of the technology Service Providers should plan to budget extra for monitoring and tools. The first networks should be instrumented down to the DSLAM/QAM and home if possible. Field tools and training for Service Provider personnel on technology is also suggested. Everyone must have a healthy comprehension for what is quality video over IP signal so troubleshooting and networks can be built with quality. Watching the TV will not be sufficient. Typical small IPTV deployments have 100 to 200 video feeds. Only gross quality errors can be caught with the eye. Tools are required for inspection and monitoring is required for catching anything 24/7.
Will monitoring cost a lot?
Yes. Instrumenting trial and first deployment networks will require more hardware. The monitoring hardware will need to also have enough technology to measure and monitor all video streams 24/7, this level of technology at this stage of the video over IP market can be expensive.
Trying to cost reduce tools for troubleshooting and monitoring before the technology is deployed risks the larger commitment. Not understanding the quality issues may kill IPTV for some service providers before it starts. Or worse, quality flaws are built into the IPTV system that will cost millions to fix. Service Providers are spending Millions and in some cases Billions of dollars on a Triple play strategy for their long-term survival, quality cannot be overlooked.
The technology is changing fast and adapting to the needs of the Service Provider, tools and monitoring will do the same but will require budget. Like all technologies, over time and through stable expanding of the deployments and services, cost reductions will occur.
Where do you monitor?
Typical IPTV deployments have about 200 streaming TV stations over multicast IP network. The first deployments should have a monitoring device at the Headend monitoring all streams entering the network. Then within the network every critical router should have a monitor device interfaced through a Tap or SPAN port to monitor all IPTV streams.
One monitoring device should be placed at each CO or Hub to monitor that the video gets to the remote locations. In trials each DSLAM and QAM may also require a monitoring unit. In deployment only critical DSLAMs and QAM may be sufficient, this is up to Service Provider. If possible monitoring in the set top box may be required.
What do you monitor?
The monitoring device should passively monitor all streams 24/7 and among other parameters, monitor each stream for the Media Delivery Index (MDI). MDI is a measure of IP cumulative Jitter and media loss and is the IP network QoE (given good video content) per stream. Alarms and thresholds need to be on a per stream basis because often Video over IP streams come from different sources and can even be different types of video such as standard definition and high definition video.
The monitor should also monitor a stream’s presence and disappearance. For multicast, the monitor can issue IGMP join/leaves to emulate a set top box and measure Zap time. Zap time is the time between changing channels on a TV. Slow networks can impact channel-changing times which monitoring can validate.
What else can be achieved?
When the network is instrumented, OSS and NMS systems should be integrated with these monitoring devices. This will expose NOC and IT engineers to the subtleties of Video over IP and the system. Experience, day one, will pay off as new visualizations and software alarm conditions help everyone involved with the Service Provider understand the system.
Employing monitoring tools from initial network turn-on will assist Network Equipment Manufactures (NEMs) to interoperate allowing quick adjust of complex QoS and Video settings within the system for a better roll-out experience.
Longer Term
The IP portions of the monitoring will merge into the routers, switches, DSLAMs and QAM devices that will reduce costs to get coverage across the network, much like RMON stats did in the emerging days of Ethernet. Monitoring equipment invested in the early phases will be used throughout network in spots where deeper monitoring may be required. Like all monitoring equipment, that hardware is designed to dig deeper into the payloads for more analysis.
Monitoring systems will be established, costs will decline and Triple Play will become a reality. Good monitoring and tools strategies will help the winners in the Triple Play race make it to the finish line.
Download
