A multiobjective Tabu framework for the optimization and evaluation of wireless systems (0907.3777v1)

Abstract: This chapter will focus on the multiobjective formulation of an optimization problem and highlight the assets of a multiobjective Tabu implementation for such problems. An illustration of a specific Multiobjective Tabu heuristic (referred to as MO Tabu in the following) will be given for 2 particular problems arising in wireless systems. The first problem addresses the planning of access points for a WLAN network with some Quality of Service requirements and the second one provides an evaluation mean to assess the performance evaluation of a wireless sensor network. The chapter will begin with an overview of multiobjective (MO) optimization featuring the definitions and concepts of the domain (e.g. Dominance, Pareto front,...) and the main MO search heuristics available so far. We will then emphasize on the definition of a problem as a multiobjective optimization problem and illustrate it by the two examples from the field of wireless networking. The next part will focus on MO Tabu, a Tabu-inspired multiobjective heuristic and describe its assets compared to other MO heuristics. The last part of the chapter will show the results obtained with this MO Tabu strategy on the 2 wireless networks related problems. Conclusion on the use of Tabu as a multiobjective heuristic will be drawn based on the results presented so far.