While most news about satellites concentrates on the very large (often over ten tons) birds, there has been enormous growth in the use of smaller (under 100 kg/220 pounds) satellites. Called nano and micro satellites the number launched increased an average of 61 percent a year from 2010 to 2015 and given current trends over 2,500 of these smaller satellites are expected to go up between 2015 and 2025. One reason these smaller satellites are so cheap is that most are built locally by people who do not specialize in satellite construction. This is related to the fact that the low cost means more risks can be taken in the design and construction of these satellites. While larger satellites have a useful life measured in years, smaller birds are built to last for a few weeks or months. The low cost and ease of construction means that more organizations worldwide can design and build satellites. All this experimentation has led to many useful advances in satellite design that have been adapted for larger satellites. At the same time these smaller satellites are increasingly going up to replace or complement larger satellites.
The earliest (1950s and early 60s) satellites were similar in size to the larger nano satellites, mainly because the available rockets back then could not put anything larger into orbit. In the late 1990s several countries in the West (especially the U.S.) began developing very small satellites again mainly because the technology had improved to the point where small was useful. The earliest of these developed by the U.S. Department of Defense were called CubeSats. That is, their volume was no more than one liter (10x10cm or 4.1x4.1 inches) and weighed no more than 1.3 kg (three pounds). The military got the idea from the increasing use of commercial nanosatellites (which weigh no more than 6.8 kg/15 pounds). The U.S. military launched its first CubeSats in 2008 (piggybacking with a larger bird). If was quickly proven that CubeSats could be used for photo or electronic surveillance, or communications. The rapid advances in communications and sensor technology in the early 21st century made it possible to build a useful reconnaissance satellite weighing less and less. A tiny satellite like this includes solar panels to provide power. A British firm pioneered this technology in the 1990s and made it possible to get scientific satellites in orbit for a fraction of the usual price. Since 2008 several hundred CubeSats have been launched and the number is increasing each year.
These tiny satellites also have the advantage of being much more difficult to be tracked from the ground. If there are successful wartime satellite attacks, then the nanosatellites can be sent up to replace the lost birds. The lightweight satellites (from CubeSats to 200 kg minisats) can be put in orbit using smaller, solid fuel, boosters. The U.S. Navy proposed using the ICBMs fired from SSBNs (ballistic missile carrying nuclear subs), which can be put to work much more quickly than the usual liquid fueled launcher. But the solid fuel ICBMs can only put a ton or more into orbit. With CubeSats and nanosatellites, this is not a problem. There are now standard nanaosatellites that are sold ready for the installation alongside the regular payload (a larger satellite).
Since the late 1990s the U.S. Air Force has been developing and launching lightweight (under half a ton) TacSat reconnaissance satellites. The first one (Tac Sat 2) was launched in 2002, while TacSat 3 went up in 2007, and was providing data to combat commanders in Afghanistan until early 2009. Troops on the ground could communicate with the TacSat, which provided photo-reconnaissance. In late 2011 the 460 kg (1,010 pound) TacSat 4 went up. Progress in this area has continued, although some of it is classified. Smaller satellites have become important for commercial uses and scientific research, both areas with much smaller budgets than the military.
The problem with these microsatellites is the cost of getting them into orbit. The cheapest way to launch these small birds is via a solid fuel ICBM (preferably one that is being retired). Even there, the launch cost is going to be about $20 million per satellite. That's why even smaller satellites became popular, because they were compact enough to be piggybacked with a larger satellite. This is becoming the most common way to launch the nano birds, keeping the cost down to under $10,000 per microsatellite. Launching dozens of them using a retired Russian ICBM has also proved economical. Thus in mid-2014 a retired Russian ICBM (a 217 ton RS-20/SS-18) launched a record 37 satellites at once (actually at 30 second intervals) after the third stage achieved the intended 630 kilometer high orbit. The RS-20 has a max satellite payload of nearly three tons. That means the average satellite weighed about 60 kg (132 pounds). Some of the 37 satellite were quite a bit heavier and most of the 37 were microsatellites (under 10 kg/22 pounds).
Russia only has a few dozen of the RS-20s left but three times as many of the smaller (106 ton) RS-18/SS-19 ICBMS for this sort of thing. These retired ICBMs have proved economical and popular for launching the growing number of highly capable but lightweight satellites. In August 2013 a RS-18 launched a South Korean 1.4 ton KOMPSAT 5 satellite that uses a radar that can detect objects and landforms as small as one meter (39 inches) across. This satellite is mainly for obtaining geographic (land and sea) information and supporting disaster response and environmental monitoring. While the satellite was new tech, the launcher was 1970s technology that was affordable and reliable because it was military surplus.
Over 550 of the RS-20s were built since the late 1960s and production ceased in the late 1980s. Most of the 300 remaining RS-20s were dismantled per disarmament agreements. About 59 remain in service and most of the remaining SS-20s may end up as satellite launchers. The RS-18 entered service in 1975 and over 500 were built by the end of the Cold War in 1991. These were basically the same technology as the SS-20, just smaller with a maximum payload of about four tons.