Dynamic mobile access gateway for heterogeneous network using analytic hierarchy process

Abstract

The recent internet and telecommunication networks are expected to be combined together in all-Internet Protocol (IP) platform. Therefore, IP mobility is important to maintain the connectivity of a mobile user (MU) when the MU roams throughout the heterogeneous networks. Common IP mobility protocols such as Mobile IPv6 (MIPv6) has issue of high handover latency. Considering this issue, Proxy Mobile IPv6 (PMIPv6) is proposed in this research as the IP mobility protocol. Even though PMIPv6 overcomes the problem of the signalling overhead in MIPv6 and reduces Layer 3 (L3) handover latency, PMIPv6 suffers from Layer 2 (L2) high handover latency which influences the total handover latency of PMIPv6. Prior to this, multi-threshold handover algorithm is proposed in this research to be implemented in Mobile Access Gateway (MAG) to reduce the L2 handover latency in PMIPv6. Multi-threshold handover algorithm considers user’s speed for the handover decision and the speed is categorized into three groups which are slow speed, medium speed and fast speed. PMIPv6 using multi-threshold handover algorithm shows up to 17% improvement compared to the PMIPv6 using dynamic handover decision method and improvement up to 99% compared to the PMIPv6 with static handover decision method. Another issue in heterogenous network is to decide the preferable network for a specific application. Thus in this research, MAG overcomes the network selection problem using Analytic Hierarchy Process (AHP). Here, five parameters are considered for the decision making, which are cell radius, data rate, applications, cost per bit and user’s speed. In short, Dynamic Mobile Access Gateway (DMAG) is developed by combining multi-threshold handover algorithm and AHP to offer seamless handover process. The performance of DMAG has been simulated using three networks which are WiFi network, 3G network and LTE network. Simulation results prove that DMAG selects the network dynamically using the five parameters compared to a method that used Dynamic Received Signal Strength. The selection reduces handover frequency for the medium speed and fast speed MU.