Multi Attribute Edge Computing Framework for Mobility Management

Abstract

Edge based Multi Decision Making Algorithm (EMDMA) is proposed. The analysis of applications of edge computing, research work done in domains like Smart Cities and Vehicular Networks has been studied. While deliberating and reviewing the research work by many authors, special emphasis had been given on the selection of parameters used for handoff process. After an exhaustive examination, a comprehensive list of parameters in different categories namely; Network, QoS and QoE had been prepared. newlineIn order to finalise the parameters, a survey from academia and industry experts had been done and results were analyzed using SPSS tool to find most relevant parameters in all these three categories. The decisive parameters considered for study are : RSSI and Network Load from Network category, Bandwidth and Packet Loss from QoS category and Cost and User Experience from QoE category. After the selection of parameters, a complete framework for network selection has been proposed. Handoff initiation and network selection are the two important aspects of proposed algorithm. A Hierarchical Fuzzy System (HFS) based solution for handoff initiation has been proposed. priority ranking of all parameters for different traffic classes like voice, video and browsing has been done for selecting a network. FAHP technique of Multi Attribute Decision Making (MADM) has been used to assign the weights to parameters. Further to select the network, two MADM methods one fuzzy based FVIKOR and other non-fuzzy based VIKOR is used. Fuzzy based network selection gave better results as compared to the non-fuzzy method. newlineThe performance of EMDMA is compared with the traditional algorithms. After simulation, average number of handoffs for proposed algorithm is reduced to 12% in case of Voice Traffic Class and 14% in Video and Browsing Traffic Class. The probability of unsuccessful handoffs has also been reduced to 0.04 in Voice and Browsing traffic class and 0.02 in Video Traffic Class.