Satellite telecoms offer global coverage and resilience. But with users increasingly expecting more on-demand services, will satellite have a significant part to play in future networks? Or will it be sidelined by the rapid growth of high capacity, low latency and adaptable terrestrial-based cloud networks?
Cloud networking is emerging as a major trend because it meets key business needs such as reduced cost, increased flexibility and faster delivery of new services and capabilities. It is simpler to operate, easier to modernise and makes applications and technology more accessible. These benefits are available to small and medium sized enterprises and startups, not just large, established companies. The impact is a greater than five-fold increase in network traffic over the next three years, with ever increasing user demand generating opportunities for telecommunications operators.
The elephant in the room when it comes to satellite services is latency. The majority of current telecommunication satellites sit in a geo-stationary orbit 36,000km away, which is the distance travelled by light in approximately 120 milliseconds. Therefore it takes a quarter of a second to send your request and another quarter of a second to receive the information back. A half second delay is not acceptable for real-time applications such as gaming, voice over internet, web-conferencing or the emerging real-time analysis of big data, such as the recently unveiled Kinesis from Amazon Web Services. These need total latencies closer to 100ms. As a result, the use of a sole satellite path is ruled out for these types of applications. However, there are many Cloud services that can utilise satellite paths where a half second delay in the initial content request is more acceptable, such as video and music streaming – which are anticipated to consist of 8 trillion hours and 107 trillion hours by 2017 respectively – or batch-type big data. Future satellite telecommunication networks may make use of lower orbits, such as the O3b network currently being deployed, in order to reduce latency by over half. However the reality is that satellite will not be able to compete with growing ground-based fibre networks – generally the shortest path is the quickest.
Satellite operators can benefit from cloud technology. A cloud-enabled satellite with data storage, processing and routing capabilities would allow the satellite to dynamically handle real-time user requests and data with only intermittent interaction with the ground. This could reduce the transit time by half again. Satellite may then have the potential to compete on transcontinental routes. This is a large step away from the traditional bent-pipe concept of a telecoms satellite, akin to the cloud concept of moving intelligence back into the network. There is even potential for operators to combine satellite resources with a multi-spacecraft network, forming a ‘space cloud'.
Potential benefits can flow in the other direction too, from satellite to cloud.
- Broadcast and multicast capability – Large amounts of data can be transmitted simultaneously to multiple users.
- Coverage – Information can be delivered at the same time to users located across continents, even where no terrestrial path such as optical fibre exists.
- Quality of service – This is largely independent of geographical location and distance from the data source.
- Reliability – High-reliability satellites can provide almost 100% availability over their lifetime.
- Resilience – Satellites offer a path independent to the ground that can be utilised in the occurrence of natural disasters.
- Homogeneity – It is easier for satellites to provide a homogenous data path.
Data-centres can be located almost anywhere on the globe. The coverage, broadcast, reliability and resilience of satellite networks can provide more flexibility to the cloud service provider (CSP) when it comes to locating its primary data-centres. This allows both capital expenditure and operating expense costs to be significantly reduced, while maintaining a highly resilient network and providing the same quality of service worldwide. Existing infrastructure can be expanded in reach, capacity and resilience. Although large amounts of bandwidth are available along extensive optical fibre routes, this is not necessarily the case when moving to the extremities of the network.
Rising customer demand will rapidly reduce available capacity. As a larger variety of operations are moved into the cloud, demand versus capacity will vary substantially – hour to hour or even second to second. The ability of the cloud to dynamically respond to user demand can be enhanced by satellites. High priority, non-real-time data can be moved to a ‘space super-highway,’ bypassing congested terrestrial routes and arriving securely at its destination. CSPs will be able to access users in areas that used to be beyond the reach of their networks. Content can be accessed via a closed and largely homogenous network from almost anywhere at any time. The satellite network, which does not rely on local terrestrial infrastructure, can provide uninterrupted access to users in the event of a disaster that cripples large areas of the ground network.
In conclusion, businesses considering moving key functions into the cloud want to know that critical applications can be accessed on demand. Integration of satellite with current and future terrestrial infrastructure could provide unprecedented levels of assurance. The cloud is already shaping a future world of telecommunications. Satellite can play an essential role within it.