Bacillus pasteurii is commonly found in wetlands, and is able to chemically create calcite. By unleashing the bacteria on areas of the desert, sand could be solidified into sandstone within a few hours. The way Larsson proposes to do this is fill massive balloons with bacteria and station them along the Sahara's southern border, where the weight of the oncoming waves of sand would pop the balloons. The released bacteria would then quickly set up a protective wall to block future sand shifts.
http://www.popsci.com/environment/article/2009-07/preventing-expansion-sahara-bacteria-built-walls
The theory of the Saharan heat pump proposes that the point of maximum sunlight, which ironically creates so much heat that it causes air to rise quickly and pulls in moisture over the mountains from the oceans, used to be over the Sahara, making it a land of Savana grasses, forests and lakes. As the tilt of the earth changed the hot zone moved south (no comments), turning the Congo from forest or whatever it was (not desert) into a wet jungle and leaving the Sahara dry.
Many proponents of this theory suspect that this will reverse itself but take a long time by human standards. I see nothing wrong with hastening this but there are some already proven desert reclamation techniques that I think are superior to building a bacteria/sand wall. Nations on the southern edge of the Sahara are already at it and it is a self-reinforcing process just as desert expansion is. More moisture causes more vegetation which captures more moisture as dew, promotes rain and moderates the temperature. Less moisture leaves less vegetation and less moderation. Push the system far enough in a new direction and it will switch equilibrium states.