Greedy randomized adaptive search procedure for analog test point selection

Abstract

Analog test point selection (ATPS) is an important problem that arises in the area of analog system testing. This paper formulates the problem as a combinatorial problem and proposes a solution method based on greedy randomized adaptive search procedure (GRASP). The proposed method is an iterative procedure, with each iteration consisting of two phases. The first phase, a construction phase, produces a feasible solution. The second, a local search, seeks for improvement on construction solution. In addition to applying the basic GRASP, the algorithm introduces randomness into both phases including: randomizing the selection of greedy criteria and checking redundant test points in a random order. The former can prevent the algorithm from converging prematurely to local optima, while the latter make the algorithm probably get more than one best solution. The efficiency of the proposed method is proven by two practical analog circuits as well as statistical experiments. Results show that our algorithm, compared with other methods, finds the global minimum set of test points more accurately and more efficiently. Therefore, it is a good solution to optimize ATPS.