Mobile ad hoc networks group mobility models of Hajj crowd dynamics

Transcription

1 Mobile ad hoc networks group mobility models of Hajj crowd dynamics Oussama Mohamed Reda Laboratoire de recherch en informatique (LRI) Department of Computer Sciences, Faculty of Sciences of Rabat, Mohammed V University in Rabat, Morocco. Abstract Hajj is a big important annual event for the Muslim world, gathering millions of pilgrims. Crowd management is one of the crucial objectives of Hajj managers. We analyze existing systems of Hajj management and show their limitations to capture its specific requirements. Unlike existing systems which put the pilgrim in the center of their models, we show that Hajj is essentially based on pilgrims being grouped together to perform their duties. We show Hajj can be modeled as a mobile ad hoc network. We propose a new approach to design and model Hajj crowd dynamics based on group mobility models of mobile ad hoc networks MANETs. INTRODUCTION Hajj is one of the five pillars of Islam. Every Muslim is required to perform Hajj at least once in a lifetime whenever it is possible for him to do it. Hajj is achieved by performing a series of actions that involve movement of pilgrims between different locations. Hajj attracts millions of Muslims every year. Those Muslims that come to perform the Hajj events are known to be called as hajjis or pilgrims. Pilgrims are due to perform specific rituals in specific times at specific locations. These spatial and temporal constraints intrinsic to the Hajj event causes huge numbers of pilgrims to be moving or standing at periods of time, together, throughout the Hajj event. It is then natural that crowd related problems arise. Frequent problems encountered during gathering events, particularly at Hajj, are amongst others, overcrowding, stampedes, overcrowding leading to stampedes or health problems, loss of pilgrims, difficulty of identifying, locating or tracking missing pilgrims, etc. It is a daunting task for Hajj officials to manage these large crowds whether they are immobile or moving between Hajj ritual places. Risks of problems leading to disasters arise when crowds are not managed or controlled efficiently. Despite many attempts, no traditional method has proven until nowadays to be efficient to address these problems with crowds. In fact, it is necessary to understand the crowds if we want to have any hope to be able to manage them. Indeed, traditional methods have never succeeded in doing this because they lacked information about crowds both on a macro level and micro level. Only with precise data or at least estimation of several parameters about the crowds can we attempt to understand and thus, manage them. Such parameters as the number or throughput of people in same or different times at same or different specific locations are highly likely impossible to be counted manually during the Hajj event. It is practically impossible to do it because of geographical, psychological and cultural constraints combined with very large moving crowds. However, technology have made, over time, huge advances in providing tools that makes it possible for pervasive and ubiquitous computing to be deployed efficiently in many areas of daily life activities. Two of these technologies are wireless networking and nanotechnology. With these two combined, it became possible for very small scale computing systems to communicate information with each other in the context of wireless networks. Radio frequency identification (RFID) represents such a technology where small-scale sized devices attached to assets of different kinds use electromagnetic waves for the identification of these assets. The use of RFID technology has then gone beyond identification, to tracking, searching or counting assets, etc. Over the last decade many works have raised interest in the benefits of using RFID technology for identification, tracking, counting or searching pilgrims during Hajj events. All these works agreed on utilizing RFID technology as a solution to help in the management of events of the magnitude of the Hajj. Indeed, [1], [2], [3], [4], [5] agreed on deploying RFID systems would assist Hajj authorities to overcome Hajj crowd management problems, such as identifying, tracking and guiding lost pilgrims at Hajj events. The RFID systems would be composed of tags and readers of these tags where every pilgrim receives a tag containing detailed information about him/her (name, age, nationality, Tawafa Establishment, group guide name and contact, etc). Hajj officials would use RFID readers to support their decision making by scanning the tags and retrieving information of contextual interest. The RFID systems would also be used for estimating pilgrim's number at Hajj ritual places areas to help Hajj staff and operators avoid overcrowding at these places. In doing so, officials would accomplish one of the crucial objectives of Hajj management which is supporting safety, security, comfort or ease for hajjis. Although these works proposed RFID systems to help ease officials burden in Hajj management tasks they remained limited in their approach to solving Hajj crowding problems specifically. In order to efficiently manage huge moving crowds in Hajj we need to understand crowd dynamics, their formation causes, their formation patterns and how they behave in specific ways. None of those works has presented a system capable of providing a platform to study and analyze dynamics of the crowds so as to understand, and thus, predict crowds 10772

2 movements. The proposed systems could be extended to be able to drive mobility models of pilgrims at different levels, yet they remain limited to efficiently and realistically comprehend crowds movements. Indeed, all the proposed systems can be extended to drive actual mobility models of individual pilgrims during the Hajj event. However, it will only reflect a narrow perspective on crowds reality. In fact, mobility models of individual pilgrims excluding their relationships to each others is not suitable to model mobility in events such as Hajj where pilgrims travel to Hajj in groups, perform their rituals in groups and stay the majority of the event within the same group. We will show throughout the rest of the paper and give evidence of why group mobility models of mobile ad hoc networks are the best approach to understand Hajj crowd dynamics. In section 2, we show why it is mandatory for every pilgrim to adhere to a group of pilgrims to be allowed to travel to perform Hajj. In section 3 we show why Hajj is naturally an event where huge crowds are constantly moving between different sites in a limited geographical environment. In section 4 we show why Hajj event can be assimilated to an ad hoc mobile network and present ad hoc mobile networks group mobility models in the literature. In section 5 we propose group mobility models of Hajj crowd dynamics. Finally, conclusions and perspectives end the paper. PILGRIMS' GROUPS OF TRAVELS AND LEADERS: THE GATE TO HAJJ Pilgrims are required to register and travel with an authorized licensed agent from their home country in order to get Hajj visa. Only authorized licensed agents hold a permit from the Saudi government authorities [6] as being legal Hajj service providers who are directly accountable to them. These authorized licensed agents specialize in forming travel groups and organizing trips to Mecca (Hajj travels) and work with local (saudi) authorities to meet their needs for transportation, accommodation, food, sleeping quarters, spiritual guidance, etc during the pilgrimage. Another type of Hajj service providers are sub-agents who merely get groups of pilgrims together to form a larger group (known as a travel group) and submit the paperwork through one of the licensed agents. Making the trip is not easy, and the Saudi Arabian Ministry of Hajj requires all Muslims perform the pilgrimage with a travel group. The only way a person can fulfill this requirement is through an authorized licensed travel agency approved by the ministry of Hajj. Travel agencies offer wide range of Hajj and Umrah travel packages for different kinds of customers along with their travel groups. Travel packages are available for individuals, couples, groups, and families. There are also customized package plans that are tailor-made to meet customers requirements and needs, whether they are individual customers, groups traveling together or families. Group packages are for those who Enjoy to perform Hajj or Umrah by being part of a group. Different group level packages are in offer. The majority of customers enjoy Family packages that provide all the services to ensure a quality stay and travel for families. Customers that have kids or parents that are accompanying them on the travel and need special care choose these kind of packages. Finally, decent number of people take up individual packages. These are people on the run and short of time or have the time but are going alone. Women over age 45 may travel without a mahram in an organized group, provided they submit a notarized letter of no objection from the son, husband or brother, authorizing travel for Hajj with the named group. Travelers found breaking this Saudi Government rule may face deportation by local immigration authorities. Travel agencies get pilgrims of different types with different travel packages (individual, couples, family, group) to constitute, in the first place, larger heterogeneous groups which represent unitary travel groups. Each travel group is designated in principle to perform the whole Hajj as a unit from their departure from their home countries, going trough the trip and performing their Hajj duties, up to arrival back to their home countries. As an essential service, approved travel agents should organize their group of pilgrims to help them, in the best possible way, fulfill comfortably and more importantly, achieve rightly one of Islam's most important requirements: the Hajj. To that end, travel agencies have different group leaders and guides that accompany the travel group for the whole trip. Among other group leaders, religious leaders (spiritual guides) have the task to help the group fulfill their obligation (Hajj duties) in a satisfactory manner. This is the most valuable service provided by travel agencies. As a first time Hajji it is natural to be unfamiliar with the rites, so, in order not for hajjis to waste their trip by missing something important, a religious leader should always be there to guide and lead his group to perform hajj duties correctly at each step. PERFORMING HAJJ: MOVING GROUPS BETWEEN RITUALS' SITES The rites of the Hajj last for five days from the 8th to the 12th of Dhu al-hijjah, the 12th month of the Islamic calendar. On the first day of Hajj, pilgrims begin their Hajj right by proceeding towards the Ka bah to perform the Tawaf which is a ritual where Muslims circle the Ka bah seven times, praying as they are doing so (figure 1; 1&2); (figure 2). When they have completed seven circles, they are finished and can now, eventuelly, begin to perform Sa ey which means to run or to make an effort. Sa ey can be delayed untill later in between the 8th and the 12th of Dhu al-hijjah. Sa ey consists of walking back and forth seven times between the hills of Safa and Marwah which are to the South and North of the Ka bah, respectively (figure 1; 3); (figure 2). Originally, this was done outdoors, but today, the entire path is enclosed in a long gallery. When pilgrims reach the top of Safa, they mark a pause time to recite specific prayers and, eventually, wait for other pilgrims of their group to rejoin them. After the pause, they proceed towards Marwah. At the top of Marwah, they pause for a certain period of time as they did in Safa, repeat their prayers or supplications while facing the Ka bah, then walk down from the hill again

3 When they finish walking back and forth seven times between safa and Marwah, pilgrims head to Mina, a town near Makkah, where they spend the rest of the day in prepared tents for the event (figure 1; 4); (figure 2). Since no major rituals take place on the first night, pilgrims stay there untill appropriate time for next duty comes which is heading to Arafat to perform Waquf (figure 1; 5); (figure 2). Indeed, on the second day of Hajj, pilgrims travel to Arafat, a nearby mountain they must reach by the afternoon, because, at this time, a ritual called Waquf begins. From the time when the sun first starts to decline until the time it sets completely, pilgrims hold a vigil on a plain of Arafat (stand in Arafat, at least for a few minutes ) during which time they pray and reflect. Hajj is Arafat. Pilgrims must stand in, in order for their Hajj to be valid. After sundown, pilgrims head to a place called Muzdalifah between Mina and Arafat (figure 1; 6); (figure 2). Here, they offer Maghrib and Isha prayers and spend the night sleeping on the ground beneath the open sky. Before the sun rises, pilgrims head back to Mina (figure 1; 7); (figure 2). Here, pilgrims participate in Ramy, a ceremony meant to symbolize stoning the devil. Pilgrims throw seven consecutive pebbles at a special stone monument called the Jamrat al Aqabah (figure 1; 8); (figure 2). This ceremony can be extremely crowded, tense, and emotional with risks of trampling deaths to occurr. Because of this, the elderly, sick, and injured are discouraged from participating. Instead, they may perform this later in the evening or have a friend perform the ritual in their place. After the Ramy ceremony, it is necessary to offer an animal sacrifice (Qurbani) to Allah. In the past, each pilgrim did this individually; however, today, it s much more common for pilgrims to simply purchase a sacrifice voucher. These vouchers signify that an animal was sacrificed in your name. After selling vouchers, qualified personnel will sacrifice a lamb for each pilgrim (or a camel for every seven pilgrims), butcher the animals, package the meat, and ship it to Muslim communities all over the world to be used to feed the poor. Animal sacrifice can be done at any point on the 10th, 11th, or 12th day of Dhu al-hijjah. If Ramy has to be postponed for any reason, pilgrims can wait until after Ramy to make their sacrifice. Pilgrims Men must have their hair ritually cut. In contrast, women can have a short lock of hair cut their heads are not shaved. The Hajj requires pilgrims to perform the Tawaf and Sa ey rituals at the Ka bah and the nearby hills. It is highly recommended that these ceremonies be done only after the stoning, sacrifice, and hair-cutting rituals. At this point Sa ey is optional depending on pilgrims having done it after the first tawaf. At the end of the third day, pilgrims return to Mina and spend the night there in prayer. Pilgrims repeat Ramy after sundown on the fourth and fifth days. In Mina, pilgrims must once again participate in the stoning ritual. This time, they will not throw pebbles only at the Jamrat al Aqabah, but also at two other monuments the Jamrat Oolah and Jamrat Wustah. First, pilgrims throw pebbles at the Jamrat Oolah, then praise and supplicate Allah. They will repeat this for the Jamrat Wustah. Finally, they will throw their pebbles at the Jamrat al Aqabah, after wich they don t need to pray they can return home. They will repeat this ritual after sundown on the fifth day. Figure 1: Ritual sites of Hajj 10774

4 Figure 2: Flowchart diagram of steps to performing the whole Hajj To mark the end of the hajj, pilgrims perform one final tawaf (Farewell Tawaf), walking around the Ka bah seven times as before (figure 1; 9); (figure 2). When they are finished, pilgrims can depart for their home as their hajj is considered complete. Anyway, foreign pilgrims are required to leave Saudi Arabia by the 10th of Muharram (the 1st month of the Islamic calendar). CLASSIFICATION OF GROUP MOBILITY MODELS OF MOBILE AD HOC NETWORKS Family and friend groups prefer to move together as a unit. If they arrive together at an event, they move around together and leave together [7], [8]. Pilgrims often travel to perform the Hajj in groups as a sign of solidarity. Pilgrims may Contact members of their local Muslim community to see if any are performing Hajj in the same period as them, if it happens to be the case, they may want to consider coordinating their trips. In addition, particularly for Hajj, all women performing Hajj should plan to travel in the company of a Mahram a close relative, husband, in-law, etc. Women over 45 can perform Hajj without a Mahram if they are part of a large group and have a notarized letter of consent from their husband. As we have seen above, the only way a pilgrim is allowed to travel to makkah to perform Hajj is by being part of a travel group of a Hajj service provider. Moreover, existing works on RFID systems for Hajj do not take into account mobility aspects of pilgrims during the event

5 They do not build upon any mobility model to propose a solid architecture for their systems. In not doing so, it is extremely difficult to build scalable and complex systems capable of intrinsic requirements capture of big events like Hajj. Eventually, existing proposals can expand to model mobility of pilgrims based on individual mobility models or entity mobility models [9]. However, we have given evidence above that Hajj is principally a religious event with an important predominant social dimension. Pilgrims arrive to Hajj in groups, perform Hajj in groups, and leave Hajj in groups. Entity mobility models cannot catch aspects of pilgrims' connectivity to each other. The only fact that pilgrims are in the same geographical area at the same period of time with the same purpose causes them to interact with each other passively or actively, whether or not they know each other. Pilgrims can be represented as mobile nodes forming a massive dense network of connected mobile nodes. Because of social dynamics, pilgrims form an infrastructureless network and eventually sub-networks where mobile nodes interact with each other without the need of any centralized controller for coordination of their communications. The network of connected mobile pilgrims can be perfectly assimilated to a mobile ad hoc network MANET [10]. In the first place, entity mobility models were almost the only type to model MANET networks. However, recently, group mobility models have emerged as the new generation mobility models for MANETs. That has been the consequence of observing mobile nodes to manifest group behavior patterns[11], [12]. Real-life scenarios manifesting characteristics of decentralized connectivity, and patterns of correlated motion between mobile nodes, like in Hajj, are the right applications that group mobility models apprehend that entity models cannot. In group mobility models, mobile nodes of the network sharing similar properties are gathered together to form a group. Essentially, nodes of a specific group are considered to be located in the same coverage area of a certain radius defined for that group. Nodes of the same group are only allowed to move in their allocated area to be considered group s members. However, all the groups are free to move to any location in the network. According to the mobility patterns of nodes in a group moving as a unit; group mobility models in literature [11 17] can be classified into two principal categories [10]. The first category is refereed to as point based group mobility (PBGM) models[10], defined in [11, 13 15] where nodes of the same group, follow as a unit, the path of the group leader. The other class of mobility models to which we refer as region based group mobility (RBGM) models [10], is defined in [12, 16, 17], where the path of group nodes is established through a sequence of regions or areas. Both PBGM and RBGM have been further classified in [10] based on the existence of interaction among group members. Indeed, models in [11, 13, 16, 17] are said to exhibit individual group member (IGM) movements because group members do not influence each other as they move only depending on their leader or targeted regions. In contrast to IGM, models in [12, 14, 15] that are said to exhibit coordinated group member (CGM) movements have their group members involved in interactions or relationships within the group, causing them to be influenced by, or correlated with each other. In this paper, we classify group mobility models in new categories that suit the context of our work. We adopt a new terminology accordingly. Indeed, Reference Point Group Mobility (RPGM) [11] model and Reference Velocity Group Mobility (RVGM) [13] model which are two important examples of PBGM with IGM group mobility models, Group Force Mobility (GFMM) [14] Model, and, Coordination Mobility (MGCM) [7] model which are PBGM with CGM model, all these models rely on the concept of the pilgrims' leader of the group. We put those three models in a class we call Pilgrims' Leader of the Group mobility model (PLGMM). On the other hand, Reference Region Group Mobility (RRGM) Model [16] and Virtual Track Group Mobility (VTGM) Model [17] classify in another category where movements of groups of pilgrims are motivated by their target destination. We call this category Pilgrims' Target destination mobility models (PTDMM). The third category of interest in this paper is of group mobility models based on dependency between members that do not pertain to the same group. In this class, Community Based Mobility Model (CMM) [12] is an important example where influences excerted on a group member stem from another member of a different group, the two groups pertaining two different travel groups in the first place. We call this class Interdependent Pilgrims' groups' members mobility models (IPGMM). HAJJ GROUP MOBILITY MODELS OF MOBILE AD HOC NETWORKS Pilgrims' Leader of the group mobility models Reference Point Group Mobility (RPGM) [11] model and Reference Velocity Group Mobility (RVGM) [13] model are two basic models of PBGM with IGM group mobility models for MANETs that can capture Hajj pilgrims motion properties and dynamics. Indeed they rely on the idea that a group is led or guided by a group leader which defines motion characteristics of the group. In RPGM [11], speed, direction and acceleration are three essential characteristics defined by the model for the group as a unit. These characteristics are entirely defined by those of the group leader. Furthermore, two motion vectors in RPGM, namely, Group motion vector (GMV) and Random motion vector (RMV) characterize the movement of the group as a whole and its members. (GMV) represents the trajectory of the group leader whereas (RMV) represents the movement of a group member (pilgrim) within the group. The sum of these two vectors determines the new position of a pilgrim at any instant. This model requires a pilgrim to stay in the defined coverage area of the group to preserve its membership to the group. In RVGM [13], which is an extension of RPGM, two velocity vectors that encapsulates the RPGM concepts of RMV and GMV are defined. The RVGM model makes use of group velocity (GV) vector of the group leader which defines the movement of the group in terms of speed and direction, while local deviation 10776

6 velocity (LDV) vector of each group member defines the amount of deviation in velocity of a pilgrim from that of the leader. The sum of these two vectors determines the movement of a group member and group leader. In both models, RPGM and RVGM, the trajectory of the group leader can be defined based on a particular entity mobility model. The RPGM and RVGM models fit well in situations where pilgrims in a group are predominantly composed of the elderly, the minor, the illetrate or pilgrims that go to Hajj for the first time that they need to be guided by a group leader through the whole event to be able to perform their rituals correctly, avoiding any misfortune of getting lost or missing a particular duty. In addition to defining velocities similarly to RVGM, GFMM [14] defines repulsion and attraction forces between nodes to avoid situations of congestion in the network as well as physical nodes collision. While pilgrims membership to a group is determined by the strength of the attraction forces between them, repulsion forces prevents from collision between each other. The strength of the force between two pilgrims is based on the distance between them and the force distributions. Repulsion and attraction forces exhibit opposite but complementary dynamics. When the repulsion force between two pilgrims is decreasing exponentially to zero (the distance between pilgrims is increasing), the model is able to change repulsion to attraction force to help maintain group pilgrims memberships. GFMM is well suited to model situations of hajj congestion when distances between all pilgrims in the Hajj decrease, thus, maintaining groups structures. Consequently, the dynamics of the groups motion is better represented by repulsion forces which are the main drive that keeps groups moving and characterize their motion. That is, pilgrims movement is verily influenced by them following their group leaders while avoiding to collide with each other within the same group or, with pilgrims of other groups. In Coordination Mobility (MGCM) [15], all groups in the network form a higher level group called main group. All groups of the main group move in coordination towards the same target destination. The main group is formed from leaders of the groups and one of those leaders is elected leader of the main group. However, group leaders may or may not know the existence of each other because they eventually stem from different travel groups. To avoid collision among the groups, velocity vectors of group leaders are used to maintain a certain distance between them. Collision only exists when the distance between two leaders is smaller than a predefined threshold. Similarly, the movement of a group member is determined by that of its leader and further influenced by its peer group members and some deviation. Coordination Mobility can be used to model the whole event of the Hajj when all pilgrims move, eventually, at the same time between ritual sites. Coordination Mobility can be further refined to larger travel groups of pilgrims that can be considered as a sub-network of the whole network, forming a two-level architecture. In this case, the whole network is divided into sub-networks on which MGCM will be applied. Situations that can solicit this bi-layered architecture are such as when a large travel group of pilgrims that find it difficult to move, as a unit, in crowded situations, is split into smaller groups capable of penetrating the crowds easily and thus moving fluently. It is worth noting that unlike situations that can rise in PTDMM, group splitting in is not imposed by the situation like in sa'ey, instead, it is a matter of choice and can be motivated by groups' logistics, for example when all or an important proportion of pilgrims is moving to a ritual site at the same time, in a crowded environment, such as heading to muzdalifah or coming back from mina. Pilgrims' Target destination mobility models Region Based Group Mobility Models (RBGM) build upon the idea of group members taking either a pre-designed or a dynamic path determined instantaneously on actual time. This path is defined as a series of zones or domains. In RBGM with IGM model, group members do not depend on each other. Only the next zone or domain, a group member is heading to, has influence on his movement. Two well-known examples of such models are Reference Region Group Mobility (RRGM) Model [16] and Virtual Track Group Mobility (VTGM) Model [17]. In RRGM, the reference zones of the sequence that form the movement path of a group are either circular or rectangular in shape. Their user-specified sizes are determined by node density and group size. In order for a group to come to the target destination, the following process repeats: A group member progresses to the next identified reference domain such that it is increasingly closer to the target destination. Before moving to the next domain, all group members should pause in the current zone for a certain amount of time. The model allows merging and splitting scenarios for groups, respectively into larger or smaller groups, which provides flexibility for the model to absorb unexpected events such as emerging of new target destinations (hair cut, etc) or formation of new groups with faster (younger pilgrims) or slower (aged pilgrims) pace. Furthermore, RRGM andvtgm are of particular interests for Hajj as they can represent frequent imposed situations during the event. Indeed, they are convenient to model larger groups of pilgrims that are sub-divided into smaller groups and likewise, smaller groups that are being grouped together -when a ritual site is reached- in order to perform the ritual all together and that no pilgrim misses it. A typical scenario of such a case is when performing sa'ey, men split away from women in the group to run on part of the way between safaa and marwah, only to wait for their women of the group to rejoin them when they finish running. Interdependent Pilgrims' groups' members based mobility models Community Based Mobility Model (CMM) [12] models interdependency between pilgrims not pertaining to the same group. It is based on social network theory and is a RBGM with CGM model. Nodes that have strong social links with each other define a group with a likelihood of nodes being co-located together, they form what is called a community. Two matrices, namely, interaction matrix (IM) and connectivity matrix represent respectively the degree of social interaction and social connectivity of any pair of nodes in the 10777

7 group. Highly connected nodes will form a community that is associated with one specific region inside the network area. The model defines social attractivity (SA) a metric-like criteria that will determine the next community region a node will move to, which is the region of the community exerting the highest SA on it. This mobility model goes well with situations where pilgrims from distinct different travel groups -who do not know each other before arriving to Hajj- have developed social relationships between them during Hajj event. As times go by, their relationships have evolved to become socially strong. When moving between ritual sites, these new developed social relationships exert a strong influence on pilgrims who consequently leave their initial groups to eventually rejoin those that have increased, in the course of time, social attractivity for them. CONCLUSION AND PERSPECTIVES Existing systems of Hajj management offer limited solutions to comprehend crowd dynamics of pilgrims. While the RFID technological aspect of those solutions is adequate to use in such a situation, it does not rely on a comprehensive model specific to Hajj. Existing systems build, by design, upon individual pilgrims being the unit of their model. We show that Hajj is a social-dimensional event where pilgrims perform their duties in groups. Furthermore, we give evidence that pilgrims grouped together is the only way for them to be allowed to perform Hajj. We show that the whole Hajj event can be assimilated to a mobile ad hoc network (MANET). We propose MANET-based group mobility models that essentially capture Hajj specific characteristics and present use case scenarios of these group mobility models. As for the future of this work, this paper lays the ground for the design of architecture of RFID systems integrating Hajj specific characteristics to build mobile ad hoc networks that capture Hajj crowd dynamics. We are working on designing architecture of such a system. REFERENCES [1] Mohandes, M 2008, An RFID-based pilgrim identification system (a pilot study), In: Optimization of Electrical and Electronic Equipment, OPTIM th international Conference on, Transilvania University of Brasov, Romania, May 2008, IEEE. [2] Yamin, M, Huang, X & Sharma, D 2009, Wireless & Sensor Technology and Crowd Management, CURIE Journal, vol. 2, no. 1, pp [3] Khan, E 2011, "An RFID-Based System for Pilgrim Management in King Abdul Aziz International Airport", Information Management, Innovation Management and Industrial Engineering (ICIII), Shenzhen, China, November [4] Al-Hashedi, A, Arshad, M, Hj Mohamed, H, Baharuddin, A "Identifying the determinants of RFID adoption intention in Hajj organizations". In: Research and Innovation in Information Systems (ICRIIS). Kuala Lumpur, Malaysia, November IEEE. [5] Al-Hashedi, A, Arshad, M, Baharudin, A & Hj Mohamed, H 2013, RFID applications in hajj management system, In: RFID-Technologies and Applications (RFID-TA), 2013 IEEE International Conference on, Johor Bahru, Malaysia, 4-5 September 2013, IEEE. [6] [7] X. Pan, C. S. Han, K. Dauber and K. H. Law. "Human and Social Behavior in Computational Modeling and Analysis of Egress," Building Future Council Doctoral Program, Las Vegas, March 16-17, It is also published in Automation in Construction, 15(4): , [8] Challenger, R., Clegg, C.W., Robinson,M.A.: Understanding crowd behaviours: Supporting evidence, 2006 [9] Camp, T., Boleng, J., & Davies, V. (2002). A survey of mobility models for ad hoc network research. Wireless Communications and Mobile Computing, 2, [10] C. Y. Aung, B. C. Seet, M. Zhang, L. F. Xie, and P. H. J. Chong, A Review of Group Mobility Models for Mobile Ad Hoc Networks, Wireless Personal Communications, vol. 85, pp , [11] Hong, X. et al. (1999) A group mobility model for ad hoc wireless networks. In Proceedings of the 2nd ACM international workshop on modeling, analysis and simulation of wireless and mobile systems (pp ). [12] Musolesi, M., & Mascolo, C. (2006). A community based mobility model for ad hoc network research. In Proceedings of the 2nd ACM/SIGMOBILE international workshop on multi-hop ad hoc networks: From theory to reality (pp ) Florence, Italy. [13] Wang, K. H., & Li, B. (2002). Group mobility and partition prediction in wireless ad-hoc networks. In IEEE international conference on communications, New York. [14] Williams, S. A., & Huang, D. (2009). Group force mobility model and its obstacle avoidance capability. Acta Astronautica, 65, [15] Wu, K., & Yu, Q. (2006) A multi-group coordination mobility model for ad hoc networks. In Proceedings military communications conference, Washington, DC. [16] Ng, J. M., & Zhang, Y. (2005). A mobility model with group partitioning for wireless ad hoc networks. In Proceedings of the third international conference on information technology and applications (Vol. 2). [17] Zhou, B., Xu, K., & Gerla, M. (2004). Group and swarm mobility models for ad hoc network scenarios using virtual tracks. In Proceedings of military communications conference, Monterey, CA