Cluster verification model for crowd management : a Hajj case study

Abstract

The fundamental advantage of WSN is the ability to deploy them in an ad hoc manner. The clustering phenomenon plays an important role in the organization of the Wireless Sensor Network (WSN), as well as dramatically affects network performance. For this reason, there has been a considerable amount of researches into ways of creating these organizational structures or clusters. Technology provides fast and advance means of transport and up-to-date information access to people that cause large crowds at the public places. This imposes a big challenge for fast crowd processing at public places such as at the airports, railway stations and checkpoints. To minimize the risk of such crowd safety and security, verification of people is necessary which caused anunwanted increment in processing time. The processing of crowd is currently done in queue form (one person at a time) because current infrastructure implementations are secluded and do not allow crowd processing in cluster form. Processing of the crowd in the queue cause long queue and long waiting time. Long waiting time cause problem for sick, old and female people (pilgrims). Our aim is to solve the above problem by designing cluster verification model, appropriate cluster formation, multi-hop cluster verification algorithm and evaluation by simulation model which processes crowd through clustering. The simulation model will use Hajj as a case study. The data collection was conducted at the King Abdul-Aziz Airport, Jeddah and in the Holy city of Makkah Almukarmah, to calculate the Hajj crowd verification time by the existing system. Cluster verification model was designed as a proposed system, architectural components and operational phases provide help to understand the functionality of the proposed model. Formation of different clusters single, split, and multiple is done to find out maximum number of verifications of the CMs in minimum time. Design of multi-hop cluster verification algorithm provides help to understand the flow of information and to write the code in any language. The simulation model was developed using open source technologies Contiki/Cooja tool, to evaluate a study of the proposed model. The results showed that the high success rate is 100% for single, split and multiple clusters when the number of CMs are 20.The worst success rates are 26%, 30% and 34% respectively, when the number of CMs are 50.The maximum 22 out of 25 CMs verified for a single cluster and gives the 88% success rate. In case of a split cluster, the maximum 27 out of 30 CMs verified with 90% success rate and in case of multiple clusters, the maximum 26 out of 30 CMs verified with 87% success rate. The optimal verification time difference is 97.6 when the number of verified CMs are 20 for single, split and multiple clusters. The maximum verification time difference is 106.1 minutes for a single cluster when the number of CMs are 25. The maximum verification time difference is 130.8 minutes for split cluster and 125.8 minutes in multiple clusters when the number of CMs are 30. It is therefore concluded that for our model, the best suitable approach is split cluster and number of CMs are from 20 to 30 that provide the optimal number of CMs, optimal success rate and optimal verification time between the existing system and proposed system. Hence, ourcluster verification model with cluster approach can benefit to minimize the verification time to process the crowd in public places.