文摘
3G Cellular networks typically consist of a group of NodeBs connected to a Radio Network Controller (RNC), and a group of RNCs to a Serving GPRS Support Node (SGSNs) as well as to a Mobile Switching Centre (MSCs). Post deployment planning of such network is to re-plan the connectivity among the above mentioned network elements with an objective to minimize the cost of operation of the network. This planning problem is traditionally solved under single homing consideration (i.e., one NodeB is homed with only one RNC). However, a single homing solution becomes ineffective when the subscriber distribution changes over time and groups of subscribers begin to show a specific diurnal pattern of their inter-MSC/SGSN mobility. One of the solutions for this problem is dual-homing where some selected NodeBs are connected to two RNCs to reduce the complex handoffs involving two different RNCs as well as two different MSCs/SGSNs. In this paper, we have mapped the dual-homing problem into a search problem and used Simulated Annealing (SA) and Tabu Search (TS) techniques to optimally select the NodeBs and RNCs to be connected. A comparison of the performances of the two meta-heuristic techniques reveals that, though both are efficient enough to produce good solutions, TS is found to be better than SA throughout.