Bandwidth requirements is a complex matter, and even more so in a centralized SaaS offering managing interactive display devices. The required transfer rate to achieve an acceptable service level will vary greatly based on many factors, including:
All bandwidth requirements listed on this page are based on the bandwidth required to provide the device management and interactive services of devices connected to the system. Any requirements added by providing external media sources playing back OTT media or accessing 3rd party services via other applications will have to be appended to the requirements listed here.
As an example: The average YouTube video bandwidth requirement for a full HD playback experience is 5mbit/s. If you intend to provide YouTube services to a site of 200 rooms, you would then need to approximate the peak usage level. Assuming the peak usage is 25% of all rooms using YouTube, you would need to provide 50(rooms) * 5mbit/s = 250mbit/s of additional bandwidth.
Managing connected clients and providing the basic user interface and services is not very bandwidth-intensive. However, a constant connection to the central systems is required, so the stability of the uplink is key. If the Internet connectivity goes down, all devices will do a full re-configuration once connectivity is restored, and this will cause a peak in bandwidth usage.
Maintaining the basic management connection requires little bandwidth, but a rich UI page might require loading several megabytes of images. Therefore, to provide an acceptable user experience, we recommend on average 200kbit/s per connected active device but always at least a 20Mbit/s uplink.
The system does not support centrally serving video content to TVs. However, the Site Agent HTTP video server supports serving video locally. Read more about the Site Agent - HTTP Video Server. Playback of video from any 3rd party source is supported but will create new bandwidth requirements as to be calculated by the installing party.
To serve video files trough the system, a locally installed video proxy device must be connected to the system. Serving video from an appliance on the local network drastically reduces the Internet bandwidth requirements in cases where one wants a video running in the main menu or similar.
The maximum number of devices in use simultaneously at any time defines your peak usage. There are many factors that come into play here, but the base would be the site occupancy levels. Once you know the occupancy levels the type of site comes into play. An airport-hotel mainly used for connecting flights might see a lot more TV usage than a spa hotel on the sea-side - even with the same occupancy.
Based on the assumption that an average site will have a peak-usage of 50% a 100 room installation would require a 20mbit/s pipe.
The bandwidth-model does not need to be linear. A huge site would not necessarily need a multi-gigabit/s uplink to server 1000+ rooms. The main concern is for the devices to remain connected and responsive. Once you cross the 500 room boundary we are therefore assuming that any further room added will only raise the requirement by 100 Kbit/s.
Peak usage factor is the peak usage in % divided by 100.
E.g. if peek usage is 50% then the peak usage factor is 50/100 = 0.5.
So the calculation becomes:
The baseline requirement is a minimum of 20 Mbps. Therefore, if any calculation results in a speed lower than 20 Mbit/s, it should be adjusted to 20 Mbit/s.
Rooms | Peak usage | minimum recommendation |
---|---|---|
100 | 25% | 20 Mbit/s |
100 | 75% | 20 Mbit/s |
500 | 50% | 50 Mbit/s |
1000 | 50% | 75 Mbit/s |
Digital signage displays serving non-interactive content will have a much more deterministic bandwidth usage pattern as compared to an interactive TV solution. The scheduled data will run as scheduled, making requests to external services at given times depending on the type of content.
A constant connection to the central systems is required, and stability of the uplink is key. If the Internet connectivity goes down, all devices will do a full re-configuration once connectivity is restored.
Unlike the interactive TVs, the signage-screens will cache most static content locally, so unless new content has been added to the system, a site-wide reload is not as resource intensive as for interactive TV.
The connected signage displays will cache all images and videos that are uploaded to the system. This means that all images and videos you upload via the management system are downloaded and stored on the display device/STB.
Running digital signage on a non-supported screen(generic HTML mode) will cause images and video to be live streamed and not cached on the client!
A template will load content dynamically from the central system. The rate of the requests are defined by the request rate set on the template items and as well as the configuration of the time-slot displaying the template. A template containing a RSS feed and a meeting-rom event item will load the contents for these items in two separate requests and the refresh rate for the content can be set separately per item.
A time-slot displaying an image and a template each at 20 seconds durations, will cause the full template to load 1.5 times every minute. That means 90 times an hour.
Let's assume the template loads 25Kb of data on each display. This will mean that this template, on this screen - and with no template item reloads, loads 2.2Mb of data an hour.
Now if this template refreshes meeting room content every 15 seconds, it will cause a reload of the data within the 20 seconds display-time. Let's assume this again causes 15Kb data to load. Now the display will load 2.7Mb an hour.
This is a reasonable configuration that can be tweaked for lower bandwidth usage if needed.
Any use of content from external systems such as playing videos from another provider, loading web pages etc. will drive bandwidth requirements up. These factors will have to be accounted for by the reseller installing the system.
Based on the above background information we reccomend a base bandwidth of 2 mbit/s + the following bandwidth-allocation per screen.
Use-case | bw per screen |
---|---|
Only images and videos served from the system | 0.25 Mbit/s |
Templates serving dynamic content | 0.5 Mbit/s |
Eg. a conference site running templated content on 10 screens will require:
All devices to be connected to the system via a stateful address-translating network device (NAT firewall) should ideally have complete outgoing access on all ports. This is our main recommendation.
If there is a specific need to restrict the number of outgoing ports that devices can connect on then the following lists can be referenced.
The IPs clients will use to communicate with the system will vary. They are all served via AWS Cloud-Front. Read more about the current IP-ranges in use by Cloud-Front here.
Note that other services set up such as OTT streams and 3rd party apps might also have a specific range of ports used for connecting to services.
The absolute ports required for professional TVs and signage displays to connect to the system are:
(all ports are outbound)
Certain manufacturers signage displays will have spesific connectivity requirements eg. to detect Internet connectivity, check for firmware versions etc.
The Mirage Site Agent needs to be able to connect to a small set of services using the following ports:
(all ports are outbound)