Towards a Dynamic Based Agents Architecture for Cellular Networks Optimisation: Cell Breathing

Abstract

One of the major optimisation issues of today’s most advanced cellular phone networks (2G, GPRS, EDGE, 3G, LTE, 4G) is the ”Network Congestion” (NC). This issue is resulting generally from an unfair distribution of the traffic between antennas. Moreover, this problem causes the increasing of dropped calls, which is unacceptable regarding today’s high standards of the mobile phone industry. To recover this traffic balance, an optimisation process is performed called ”Load Balancing” (LB). Most of the works done in this area focus more on the efficiency of the optimisation techniques rather than the dynamic and automatic aspects of these last ones. Knowing that radio telephony networks are real life applications, makes that the real time and the dynamical resolving as much important as the efficiency of the techniques used. That’s why, in this paper we have tackled for the first time the network congestion issue from a software engineering point of view. A high level modeling based on a multi-agent system and algerbra process Π-calculus is used in order to design self-adaptative, dynamical system that can respond and cope wih the network congestion issue.