The UMTS radio network planning problem poses the challenge of designing a cost-effective network that provides users with sufficient coverage and capacity. We describe an optimization model for this problem that is based on comprehensive planning data of the EU project MOMENTUM. We present heuristic mathematical methods for this realistic model, including computational results.
Wireless Local Area Network (WLAN) is currently among the most important technologies for wireless broadband
access. The IEEE 802.11 technology is attractive for its maturity and low equipment costs. The overall performance of a specific
WLAN installation is largely determined by the network layout and the radio channels used. Optimizing these design parameters
can greatly improve performance.
In this paper, access point (AP) placement and channel assignment is optimized using mathematical programming. Traditionally,
these decisions are taken sequentially; AP placement is often modeled as a facility location problem, channel assignment
as an (extended) graph coloring problem. Treating these key decisions separately may lead to suboptimal designs. We propose
an integrated model that addresses both aspects. The different optimization objectives and their tradeoff are taken into consideration simultaneously. Computational results show that indeed the integrated approach is superior to the sequential one.
UMTS radio network evaluation and design are currently important issues for telecommunication operators.
We present a novel view on network evaluation. The recent dimension reduction approach is
generalized to an analytical approximation of the network's general performance based on average traffic.
The pivot is an average coupling matrix that captures the essential coverage and cell coupling properties of
the radio network. Based on this new evaluation method, we present new optimization methods, namely
a new optimization model based on designing the generalized average coupling matrix and an efficient
1-opt local search. We give preliminary computational results that show the potential of our methods on
realistic data.