Satellite Internet connections are widely used these days, but there still are limitations to these services, due to some technical reasons. Signal latency is one of them. It is the delay that appears between the moment when data is requested and the moment when the response is received. When the communication is in one direction only, this is the time passed between the moment of the signal’s broadcast and the time when it is received at destination.
The magnitude of the latency depends on the distance travelled, as well as on the speed of light. About 0.12 seconds are necessary for the wireless radiation to cover the distance of 36,000 km to a geostationary satellite positioned above the equator. As a consequence, a complete session would take one quarter of a second. Latency is the difference between standard terrestrial based networks and geostationary satellite networks. Geostationary satellite communications networks have a twenty times bigger latency than terrestrial based networks, according to NetworkReviews.net.
Geostationary Satellites and Low-Latency Applications
Geostationary orbits (GEO – geostationary Earth orbits) are geosynchronous orbits above the equator (0º latitude). They have a period equal to the rotational period of the Earth, while the eccentricity of the orbit is practically zero, i.e. the orbit is circular. A satellite in a geostationary orbit seems motionless to ground observers. It seems to have a fixed position when they look at it in the sky. This is often the case of weather and communications satellites. As a result, satellite antennas that target them do not have to change their position, since they can permanently point the same position. The satellites used for satellite Internet access only differ by longitude, as they are all positioned above the Earth’s equator.
The higher latency of geostationary satellites is due to the 35,786 km that the signal has to travel two times: to the satellite and back to Earth. It does not matter that the signal travels at 300,000 km/s, the speed of light, because the delay still remains significant. Even if any other delays can be eliminated, the radio signal will still take one quarter of a second to reach the satellite and come back again to the ground. This delay varies depending on the position of the users – right below the satellite or far, near the horizon.
For Internet packets, the delay is doubled before the reply is received. The connection latency for satellite Internet connections is 500-700 ms from a user to the Internet service provider (ISP), one way only, while the total RTT (round-trip time) is between 1,000 and 1,400 ms. Compared to dial-up connections, where users experience 150-200 ms typical latency, and 15-40 ms latency in the case of high-speed internet services (cable or VDSL), satellite Internet connections are slower.
It is not possible to eliminate latency in the case of geostationary satellites. However, this problem can be mitigated in communications with TCP acceleration, which can shorten the RTT per packet by using a technique that allows a split of the feedback loop between the receiver and the sender. Recent technology developments included in satellite Internet services provide such acceleration features.
Latency also has negative effects on SSL and other secure Internet connections which require numerous data exchanges between the web server and the client. Although small pieces of data are involved, the big number of round trips produces significant delays, unlike other internet connectivity methods. The effect is visible when applications like data entering or editing are used, as well as other types of online work. Live interactive access to distant computers, such as virtual private networks, for instance, works better with the newest generation of satellite Internet service.