Multilevel refinement for combinatorial optimisation: boosting metaheuristic performance

The multilevel paradigm as applied to combinatorial optimisation problems is a simple one, which at its most basic involves recursive coarsening to create a hierarchy of approximations to the original problem. An initial solution is found, usually at the coarsest level, and then iteratively refined at each level, coarsest to finest, typically by using some kind of heuristic optimisation algorithm (either a problem-specific local search scheme or a metaheuristic). Solution extension (or projection) operators can transfer the solution from one level to another. As a general solution strategy, the multilevel paradigm has been in use for many years and has been applied to many problem areas (for example multigrid techniques can be viewed as a prime example of the paradigm). Overview papers such as [] attest to its efficacy. However, with the exception of the graph partitioning problem, multilevel techniques have not been widely applied to combinatorial problems and in this chapter we discuss recent developments. In this chapter we survey the use of multilevel combinatorial techniques and consider their ability to boost the performance of (meta)heuristic optimisation algorithms.

Personal and proprietary tenancies

Examines the Chancery Division ruling in London Development Agency v Nidai on whether a number of agreements providing for the construction of a bridge and shop premises on the retaining walls of a river resulted in a binding legal lease or a series of bare licences. Comments on the failure of the judgment to mention the House of Lords ruling in Bruton v London & Quadrant Housing Trust and discusses whether a Bruton tenancy is capable of binding third parties.

Megasonic agitation for enhanced electrodeposition of copper

In this paper we propose an agitation method based on megasonic acoustic streaming to overcome the limitations in plating rate and uniformity of the metal deposits during the electroplating process. Megasonic agitation at a frequency of 1 MHz allows the reduction of the thickness of the Nernst diffusion layer to less than 600 nm. Two applications that demonstrate the benefits of megasonic acoustic streaming are presented: the formation of uniform ultra-fine pitch flip-chip bumps and the metallisation of high aspect ratio microvias. For the latter application, a multi-physics based numerical simulation is implemented to describe the hydrodynamics introduced by the acoustic waves as they travel inside the deep microvias.