1: \begin{abstract}This paper considers a novel problem of how to

2: choose an appropriate geometry for a group of agents with only shape

3: constraints but with a flexible scale. Instead of assigning the

4: formation system with a specific geometry, here the only requirement

5: on the desired geometry is a shape without any location, rotation

6: and, most importantly, scale constraints. Optimal rigid

7: transformation between two different geometries is discussed with

8: especial focus on the scaling operation, and the cooperative

9: performance of the system is evaluated by what we call the geometries

10:  \emph{degrees of similarity} (DOS) with respect to the desired shape during the entire convergence process. The design  of the

11: scale when measuring the DOS is discussed from constant value and time-varying function

12: perspectives respectively. Fixed structured nonlinear  control laws

13: that are functions on the scale are developed to guarantee the

14: exponential convergence of the system to the assigned shape. Our

15: research is originated from a three-agent formation system and is

16: further extended to multiple ($n>3$) agents by defining a

17: \emph{triangular complement graph}. Simulations demonstrate that

18: formation system with the time-varying scale function outperforms

19: the one with an arbitrary constant scale, and the relationship

20: between underlying topology and the system performance is further

21: discussed based on the simulation observations. Moveover, the

22: control scheme is applied to bearing-only sensor-target localization

23: to show its application potentials.

24: \end{abstract}