Over the last 30 years there have been a number of projects to create a satellite constellation, low orbit, able to create a communication infrastructure in the sky. Back at the end of the 80ies the idea of a constellation of satellites turned into a concrete project with Iridium, a cooperative effort involving many companies and led by Motorola. Iridium consists of 66 active satellites plus a few serving as spare and orbiting at an altitude (average) of 781 km. It became operational in 1998 providing voice, messages and data services.
Starlink is the last in a row of constellation projects and it benefits from new launching infrastructure and leading edge electronics keeping the size and cost of satellites low.
Once completed the first phase, end of 2020, it will consist of 1584 satellites organised in 72 orbits each one with 22 satellites at an average altitude of 550 km. With the launch on Aprils 22nd, 2020, with the Falcon 9, 60 more satellites have been put into a low 220km orbit where they will remain for testing and then will be moved to their final orbit position, 550 km above the Earth. As of June 2020 the first 422 satellites are in place, a bit more than 1/4 of the final target. These first 422 satellites are sufficient to start servicing the nordic parts of the planet. 2 further launches in May and June have brought the number up to 540 satellites in orbit (but only 422 operational at the end of June)
Each satellites is some 260kg, much lighter than those first Iridium satellites (689kg). Once all satellites will be in orbit, end of 2020, they should be able to deliver broadband connectivity (in the hundreds of Mbps range) to most planet populated areas. However, the full coverage able to deliver dense connectivity will have to wait for phase two when some 30,000 satellites will be in orbit.
The connectivity infrastructure in the sky is used through ground transponders each the size of a pizza box (that can be solar powered) and have the required hardware to track the satellites with their antennas. The user terminal can be connected to the transponders in a variety of ways, including WiFi. Hence, the Starlink system does not connect directly like Iridium or Inmarsat, Globalstar to a satellite cellphone. The reason is that these satellite phones are expensive and their capability to track the satellite signal is limited requiring higher transmission power. With the ground transponder any normal smartphone will be able to connect.
This infrastructure, some claim, can have a disruptive impact on Telecom Operators once fully completed. Basically it provides an alternative to land infrastructures.
Starlink is by no means the arrival point for satellite constellations. Researchers are working on nano-satellites, also known as cubsats since they look like tiny cubes, that will be very cheap to build (scale factor at play) and to put in orbit (very light).
Personally, I don’t think that the effect will be disruptive in the short term but it will surely provide an alternative to established infrastructures. Add to that the evolution of wireless networks where the 6G is likely to be created bottom up, through devices and terminals rather than top down by Operators and you see a new communication paradigm taking shape: from a communication infrastructure owned by a few players to a communication fabric created by users. As a matter of fact and for a long time, the two will coexist but this will be sufficient to change the telecom business forever.