Each of these networks is completely different, ranging from high-throughput with effectively no power constraints, through to very low-throughput with power measured in micro-watts. Clearly, some of these problems are harder than others.
To my mind, ‘Mobile Ad-Hoc Sensor Networks’ are the hardest of these problems, they take something which is already very difficult, an ‘Ad-Hoc Sensor Network’ and add an uncertainty of position and network topology, caused by physical mobility of the nodes. Let’s start by considering ‘normal’ Ad-Hoc Sensor Networks, these are normally very power constrained systems, examples of applications would be:
Usually the sensors must be distributed, and in some use cases there may be hundreds or thousands of them, and so as a simple matter of cost they have a requirement to survive several years before they are serviced. The large number of them, and area covered, tends to make providing infrastructure cabling or repeating base stations rather unappealing due to the time, materials and cost involved (imagine trying to place a WiFi base station every 100m across an entire farm).
As a result, these systems typically rely on mesh networks, mainly for their ability to work without a fixed infrastructure, but another frequently quoted reason is that the mesh topology reduces the distance that each transmission must travel. This allows for substantial power saving, as radio power and distance are related by the inverse-square law (half the distance theoretically requires a quarter of the power).
For More Details: 60-90 Seconds Video Cost