Ordered Escape Routing via Assignment of Routability-Driven Detouring Paths

Abstract

In board-level routing, ordered escape routing becomes a key problem. In this paper, given a set of escape pins inside a pin array, a set of available boundaries and the capacity constraint between two adjacent pins, an efficient algorithm can be proposed to solve the routing problem in ordered escape routing under capacity constraints. Firstly, based on the mapping process of the detouring paths of the untangled nets in one-sided net untangling, the routability-driven pins of the untangled nets can be assigned onto feasible positions and the target pins of the escape nets can be assigned on some available boundaries. Furthermore, based on the assignment of the routability-driven detouring paths with the capacity consideration, the global wires of the escape nets in a pin array can be assigned in single-layer global routing and the routing paths of the escape nets in a pin array can be assigned in single-layer detailed routing. Compared with Luo’s SAT-based algorithm, Jiao’s flow-based algorithm and Yan’s heuristic algorithm for the 6 tested examples with capacity as 1, our proposed algorithm can obtain the same 100% routability of the escape nets. Compared with Yan’s heuristic algorithm for the other 6 tested examples with capacity as 2, our proposed algorithm can use reasonable CPU time to improve 1.9% of the total length of the routed escape nets and 0.25% of routability of the routed escape nets on the achievement of the 100% routability on the average.