|
|
 |
| |
|
|
|
LWDF-DiscMOPSO-EDF Hybrid OFDM Downlink Schedulers for Improved Performance in 4G Networks |
|
|
|
PP: 73-95 |
|
|
|
doi:10.18576/amis/13S108
|
|
|
|
Author(s) |
|
|
|
T. Sentthil Vinayakam,
A. Rajeswari,
|
|
|
|
Abstract |
|
|
| A hybrid scheduling scheme for the downlink of 4G wireless network is proposed herewith and it provides improved and consistent performance across all performance metrics for various overloaded traffic conditions with various heterogeneous traffic mix ratios and intensities. The proposed hybrid algorithm, namely ”LWDF-DiscMOPSO-EDF Hybrid (LDEH)” OFDM scheduling scheme is an innovative and improvised combination of the best logic and capabilities of multiple scheduling algorithms i.e. PSO, LWDF and EDF. The PSO-based heuristic scheduler has been improvised much into discrete-multi-objective PSO that is designed for achieving multiple objectives, namely best goodput, reduced starvation, deadline and PLR compliances, all targeted simultaneously. The proposed discrete-PSO overcomes the high spectral inefficiency of a conventional PSO. The well-considered design of the bandwidth-reservation policy in this LDEH scheme enables compliance with deadline & other QoS requirements even under high-intensity traffic conditions. Two versions of the LDEH hybrid scheme are proposed based on two different bandwidth reservation policies. These hybrid schemes’ performance is compared with various contemporary algorithms and results prove that both the proposed LDEH schemes outperform existing schemes by producing the best overall and consistent rankings across various metrics and classes of service. The proposed schedulers are also fair, i.e. they do not compromise on performance for one metric or class to improve in another metric or class. Demanding traffic scenarios are used for simulation so that they tested the schedulers in various over-loaded traffic conditions and with varied traffic mix ratios. |
|
|
|
|
|
|
|
