Ensuring high-signal quality and availability is as indispensable in digital TV networks as it was with analog networks. Signal monitoring is key to quality assurance, but it is getting increasingly complex — not least because of the wide variety of programs and auxiliary data services. This article covers where to monitor a network, the measurements that need to be performed and the requirements for the monitoring instruments.
Creating the perfect monitoring solution
Designing the ideal system for monitoring digital TV networks will be different for each application. It will depend on the objectives, the network structure — at which point in the network is the monitoring to be performed, what signals are to be monitored, and what measurements will need to be performed.
Program providers and network operators place considerable importance on the error-free generation of digital TV signals and their correct distribution and transmission over networks, not least because of contractual obligations toward their business partners. Government regulators undertake different tasks. They check if relevant standards are complied with, including whether the coverage in a specific transmission area is adequate, or if a specific program multiplex is broadcast correctly.
There is a wide variety of monitoring objectives, including:
- preventing serious interference;
- immediate and efficient troubleshooting by quickly detecting and locating signal failures or signal errors and determining their causes;
- the recording of signal characteristics and system availability for subsequent analysis or demonstration purposes toward the contracting partners; and
- enhancing quality of service by detecting and remedying even sporadic errors.
Networks for distributing digital TV signals are often complex. For example, network segments or feeder links are frequently operated by different organizations. A variety of methods are used to physically transmit signals. Cable headends, for example, can be supplied via satellites or fiber-optic links (IP or ATM), while standalone terrestrial transmitters can be connected via microwave links or off-air reception, and terrestrial multi-frequency networks (MFNs) may also include local single-frequency network (SFN) sections. (See Figure 1.)
Typical monitoring points in the network
The interface to the customer is very important in monitoring; this is where the signal is checked for errors. Monitoring of the transmitter modulation quality must be carried out directly on the transmitter. The SFN characteristics must be monitored at a point where the reception of all transmitters included in the SFN is good. If transport streams are combined or modified, the newly generated transport stream should be monitored directly on-site to ensure that it is correct. If the individual network sections are the responsibility of different service providers, monitoring should be performed at the transfer points between two providers. The individual programs are usually monitored directly at the site of program provision or channel coding.
Which measurements need to be performed?
If the transport stream is to be distributed without any changes, for example via a terrestrial transmitter network, monitoring will focus on the RF characteristics. If, however, the network operator modifies the transport stream, additional monitoring of the transport stream characteristics can be beneficial.
But program providers mainly concentrate on the video and audio signal quality and the correct structure of the transport stream.
Measurements from the signal processing chain
Figure 2 shows the digital TV signal processing chain in simplified form. The signals may be sourced from a live feed or a server. The multiplexer combines the individual, compressed signals and adds MPEG service information. In addition to the audio and video signals, all other data such as Teletext, subtitles and program information for an EPG is inserted. The multiplexer output is a transport stream. Depending on the transmission method, the transport stream is channel-coded, modulated and then transmitted terrestrially via satellite uplink or through a cable network. Based on the signal processing chain in Figure 2, the following measurements can be made:
- program level; encoder output (point 1) — poor picture and sound quality, coding syntax errors, incorrect time references (e.g. program clock reference and presentation and decode time stamps) and unnecessarily high data rate;
- transport stream level; multiplexer output (point 2) — incorrect or missing elementary streams, data rate too high or too low, incorrect references and syntax errors; and
- RF signal level (point 3) — missing channel, insufficient signal strength/quality, or BER too high, and impaired transmitter synchronism in SFNs.
Recommendations of the measurement guidelines
The DVB Project has published a technical report, ETSI TR 101 290, titled “Measurement guidelines for DVB systems.” This report specifies how to perform measurements on DVB systems both with regard to RF and transport stream characteristics. It also suggests measurements especially with a view to transport stream monitoring. The measurements are classified in three groups and prioritized. All measurements refer to the syntax and the logical structure (references in tables) of the transport stream as well as to the characteristics with time reference (PCR and buffer) and the integrity (CRC).
In practical applications, the measurements focus on RF and transport stream characteristics. In some cases, the individual programs are automatically monitored in order to check their video and audio quality, and — insofar as possible — the correct structures for the data services.
Monitoring RF characteristics
Only a few RF measurements are needed to detect the errors at the RF level as described in the DVB report. They are part of the basic measurements in practical monitoring applications: RF sync, RF level, and the modulation, bit and packet errors.
During DVB-T transmission in SFNs, transmitter synchronism must also be monitored by measuring changes in frequency, level and time reference of all transmitters in the SFN. Thus, an individual transmitter that is no longer in sync can be immediately detected and switched off. This prevents a failure of the entire network, and the deactivation of the defective transmitter affects only a few customers.
Detecting small changes of a transmitter signal at an early stage is of vital importance in RF monitoring. By taking measures early on, extensive changes and even transmitter downtimes can be prevented.
Monitoring the transport stream
By performing the measurements of priority 1, 2 and 3 as specified in the DVB measurement guidelines, the transport stream syntax and structure can be appropriately determined. In most cases, however, further measurements need to be performed.
For example, the megaframe initialization packet (MIP) is checked in SFNs. The MIP is a transport stream packet required for synchronizing the transmitters involved. A MIP inserter inserts the TS packet in a transport stream that is sent to all transmitters.
Further measurements provide answers to the following questions:
- Are all the programs available?
- Do the programs contain all desired elements?
- Does a program take up too much bandwidth?
- Are all the program designations correct?
- Are pay-TV programs encrypted?
Monitoring of programs and services
When the quality of video and audio signals is monitored, the test equipment flags picture freeze or loss as well as silence or sound loss. It is also possible to calculate a quality value for the video signal. Analysis functions and equipment are used to check the configuration of data services such as Teletext, subtitles or MHP.
Configuration: maximum flexibility and capability
For monitoring to be considered efficient, real errors must be detected but no false alarms triggered. However, monitoring tasks vary greatly, and the definitions of errors or false alarms are not necessarily standardized. To make the interpretation of measurement results easier, it is useful to be able to classify the individual measurements (e.g. as “Alarm,” “Warning,” or “Info”). This classification can then be used by the monitoring instrument for all further signaling options (e.g. for class-specific icons on the graphical user interface, filter criteria in SNMP traps, and explanations in reports).
It is useful to configure in-depth at the transport stream level, for example when a network operator transmits additional data in unreferenced transport stream packets with known Packet IDs (PIDs). These PIDs are not supposed to provoke the error message “unreferenced PID.” But the monitoring instrument must indicate other unreferenced PIDs in the transport stream as erroneous. The situation is similar when the transport streams transmitted by a network operator include known, sporadic errors that are to be ignored as long as they only last for a specified period of time. Otherwise, too many alarms would be triggered.
If you do not have a budget to buy monitoring systems that can monitor all signals simultaneously, you can monitor the signals sequentially using one measurement and demodulator unit. To provide proof for clients, errors detected by the monitoring instrument must be recorded and automatically archived to help ensure a detailed analysis on demand.
In some cases, monitoring instruments may be located at unattended stations or stations that are difficult to access. If multiple, spatially separated monitoring points are involved, the best solution is to route all measurement results to a central PC via a network interface.
Integration into network management systems
SNMP is used as standard to integrate the monitoring instrument into network management systems. This protocol enables the operator to read and write individual variables in the monitoring instrument, and thus query measurement results and modify configurations. Errors detected by the instrument can be sent as traps. This function is used to notify you at a remote location and to trigger an alarm, if required.
Monitoring the transmission and distribution of digital TV signals is a complex task. When the specifications for a monitoring system are being defined, the monitoring objectives as well as the function and structure of the network to be monitored are the key aspects. The more measurement points and the more complex and detailed the measurements, the better the information about signal characteristics, signal errors and their cause, and, likewise, the more specific and faster the response to alarms.
A monitoring system is a compromise between budget and the number of measurement points. The monitoring instrument must provide the required monitoring functions as well as simple and flexible configuration options to meet the specific requirements of the signals to be monitored.
Thomas Tobergte is product manager for broadcast test and measurement products at Rohde & Schwarz.