End- to- end QoS provisioing is employed in through various networkmanagement mechnishms involving Load balancing, Resource provisioing, Mobilitymanagement, and power manaegemnt techniques. In the first module, loadbalancing is performed based on theBandwidth Exploitation Factor (BEF), modelled as a factor of the load factor ofthe inerworking networks. The uplik and the downlink load factor of UMB, WiMAXand WLAN network are computed based on the channel conditions, such as SINR,Bit Error Rate (BER), Transmission Power, interfernce powers, and the channelbussyeness ratio, that represents the dynamic channel conditions. The Utilizedamount of Bandwidth by the MN under the Network are computed depending on thedynamic data rate and the Bandwith availabilty. The load balancing among thehybrid coupled UMB-WiMAX-WLAN interworking network is performed by the centralLaod Balancer (LB) node that redirects the traffic to the network with theoptimum load condition based on a predefined threshold value.
The developed algorithm proved to reduce the latency metrics of thereal time and non-real time applicationsobtained form the simulation. High utilization is achieved by configuring theload balancer to dynamically forward the traffic load to the network with minimumload, and by dynamically broadcastingthe network condition at the time of the client request, instead ofbroadcasting messages at regular intervals. The proposed Hybrid coupledUMB-WiMAX-WLAN- IIP architecture withLB, shows an reduced responce times and latency metirc for compared to theinternetworking scenaios without LB and Intergrated with existing IMSarchitecture. In Module 2, an Active Resource Allocation (ARA) algorithm for RadioResource management for efficient utilization of the available radio resourceamong the interworking heterogeneous wireless network has been proposed. In this deveoped ARA scheme, the availablechannels are allocated through complete partitioning to Real Time (RT) and Non-Real Time (NRT) service classes, thus providing channels for all type ofapplications.
The simultion results for Voice jitter, packet delay variation, MeanOpinion Score (MOS), video end to endpacket delay, Response times of FTP, Email and HTTP were obtained and comparedfor the developed scenarios. Comparing all the scenarios, it can be observedthe hybrid coupled UMB-WiMAX-WLAN internetworking scenario with the propsed ARA algorithm outcomes the performance of thenetwork without resource allocation mechanism in both IIP and IMS interworkingnetworks. Thus the proposed ARA scheme provides an efficient utilization ofnetwork resource with guaranteed end- to- end QoS for various service classes,by reducing the latency and responsetimes. A Multiple Decision based Mobility managementalgorithm, that engage both user and network based decision metrics isdeveloped for the hybrid coupled UMB-WiMAX-WLAN internetworking architectureover a proposed Enhanced Hybrid Mobile IP (EHMIP), in the next module. The EHMIP is developed to overcome thetriangular routing problem of the exiting PMIPV6 mobility management protocol,and to reduce the latency of signalling between nodes involved in mobility. Tomeet the QoS satisfaction of the conversational and interactive services, aMultiple Decision Mobility Management algorithm that involved both user andnetwork metrics, such as the location, mobility of user and load and type oftraffic served by network is developed to perform the handoff decision andnetwork selection. Toobtain an increased capacity, and to maximise the user experience withminimised latency, that meets therequirement of multimedia applications of the future generation technology,managing the interference power levels of BSs forms an key factor. EmployingCell Rang Expansion technique, by deploying low power Pico BS nodes along with existing Macro BSs, can result for an increased capacity of interworking networkwith increased user experience even at the cell edges.
To minimize the responsetimes and latency of various traffic type services to the value less thanmillisecond, a Smart Application based policy Driven Cell Selection (SAPDCS)algorithm is proposed in this module. The decision entity employees an applicationtraffic thresholding mechanism, that computes the response times ofapplications, and the node (either Macro or Pico BS) with minimum value thanthe predefined threshold, is selected for serving the MN. By employing thePolicy Driven based decision entity servers in the SAPDCS, the QoS measures ofnetwork is maintained by reducing the delay metrics of the traffic applicationsinvolved. An analysis of the QoS Provisioning techniques employed at eachmodules is performed for the proposedUMB-WiMAX-WLAN-IIP, UMB-WiMAX-WLAN-IMS and the existing UMTS-WiMAX-WLAN.
The investigations on the comparisonresults of the various metrics of multimedia and interactive applications ,proves the efficiency of the proposed QoS provision framework, in obtaining anoptimum delay and response time, that satisfies the future generation wirelesstechnology requirements.