Activity and Mobility Modeling of Theme Park Guests
Guest services offered in parks and resorts may rely on wireless communication in the future. Examples are location-based applications, mobile games, information guides, multimedia sharing, and mobile social networks. Rolling out a wireless infrastructure to support such services in parks may create challenges beyond deployment and maintenance. For example, access points and antennas may be too visible and might interfere with artistic intentions. However, a continuous connectivity provided by base stations and cellular infrastructure is not always needed. Instead, direct links between the wireless devices carried by guests might provide enough capacity. The adhoc mode of Wi-Fi, Bluetooth, or ZigBee enables wireless devices to communicate directly when within each other’s coverage range. Additional mesh networks might assist only where needed at key locations. Naturally, such adhoc communication will experience occasional disruptions as wireless links will appear and disappear with the guest mobility. The effect of the disruptions depends on the number and density of devices and the guest activity like walking speed or directions. Therefore, it is essential to evaluate the performance of future wireless communication with tools using realistic mobility and activity models.
Existing models and tools are either simplistic and do not correspond to specific real-world mobility scenarios, or they target common locations such as mobility in city centers. We aim to create activity models that target entertainment theme parks and capture mobility characteristics of park guests, such as their arrival and departure times and positions in the park. Our models are implemented in a simulator and used to generate realistic mobility patterns for large-scale evaluation of wireless systems. The mobility in the simulator is driven by the activities that guests are interested in while enjoying the parks (visiting attractions and street performances, restaurants, or shopping areas, etc.). We use park data and GPS traces to calibrate and validate our models.