Abstract
QoS Optimization is an important part of LTE SON, but not yet defined in the specification. We discuss modeling the problem of QoS optimization, improve the fitness function, then provide an algorithm based on MPSO to search the optimal QoS parameter value set for LTE networks. Simulation results show that the algorithm converges more quickly and more accurately than the GA which can be applied in LTE SON.
Highlights
Spectral efficiency has been greatly improved in LTE networks, but it still cannot carry the explosive growth of service data volume
We discuss modeling the problem of Quality of Services (QoS) optimization, improve the fitness function, provide an algorithm based on Multi-Level Particle Swarm Optimization (MPSO) to search the optimal QoS parameter value set for LTE networks
1) Problem coding The position of the i-th particle is described by vector Xi, the d-th element Xid of the vector represents a value of QoS parameter i, the vector magnitude is equal to the number of QoS parameters
Summary
Spectral efficiency has been greatly improved in LTE networks, but it still cannot carry the explosive growth of service data volume. Faced with this situation, the mobile communication operators decide to cancel “unlimited package”, and launch “traffic management” to improve the revenue per bit. The optimization of QoS is to find the optimal parameter set for radio resource management (RRM) which can maximize network spectral efficiency when customer experiences are met. Reference [7] applied genetic approach (GA) to QoS optimization for WCDMA mobile networks, which provided a good reference, but this method was not related with SON. The algorithm is expected to be applied in LTE SON QoS optimization. The paper is organized as follows: Section 2 addresses the problem and some conceptions, Section 3 provides the algorithm and realization, Section 4 is simulation and analysis, and Section 5 concludes the paper
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
