Density and location optimization of nomadic nodes in heterogeneous networks.

Abstract

Data traffic demand has been increasing exponentially and this trend will continue over the foreseeable future. This has forced operators to upgrade and densify their mobile networks to enhance their capacity. Future networks will be characterized by a dense deployment of small BSs. However, the fixed deployment doesnot cater for the location variant scenarios and therefore, a need for a flexible network structure through deploying Nomadic Nodes (NNs). NNs can be defined as small low power consuming BSs on wheels. NNs enable flexible network deployment hence improving capacity and coverage on demand.

We therefore propose a HetNet system model comprising of macro BSs and NNs and an optimisation framework to optimise their density and locations. The framework minimizes the number of NNs required by the network to attain the set network requirements. Thus, the power consumed by the network is minimised since it directly depends on the number of NNs deployed in the network. The proposed network was simulated in MATLAB and its performance analysed using key performance indicators such as sum rate, average user rate, probability of connectivity and energy efficiency. The performance of the optimal network was analysed in comparison to the non-optimal network.

Considering the optimal network, probability of connectivity, average user rate and sum rate initially increase as NNs are added to the network. However, they saturate at a certain point in the network because, at this point of time, all the users have been served and extra NNs carry no traffic. More NNs are required for the non-optimal network to attain the same connectivity, average user rate and sum rate. The peak EE for the optimal network is higher than that of the non-optimal network. Therefore, the optimal network is more energy efficient.