Empirical modeling of campus-wide pedestrian mobility: observations on the USC campus

Kotz

Proceedings IEEE INFOCOM 2006. 25TH IEEE International Conference on Computer Communications

2006

442

289

Abstract-Understanding user mobility is critical for simulations of mobile devices in a wireless network, but current mobility models often do not reflect real user movements. In this paper, we provide a foundation for such work by exploring mobility characteristics in traces of mobile users. We present a method to estimate the physical location of users from a large trace of mobile devices associating with access points in a wireless network. Using this method, we extracted tracks of always-on Wi-Fi devices from a 13-month trace. We discovered that the speed and pause time each follow a log-normal distribution and that the direction of movements closely reflects the direction of roads and walkways. Based on the extracted mobility characteristics, we developed a mobility model, focusing on movements among popular regions. Our validation shows that synthetic tracks match real tracks with a median relative error of 17%.

Section: Related Workmentioning

confidence: 99%

Extracting a Mobility Model from Real User Traces

Kotz

Proceedings IEEE INFOCOM 2006. 25TH IEEE International Conference on Computer Communications

2006

442

289

“…For example, [61] and [62] discuss an empirical model based on observations of pedestrians on a university campus, while [63] describes a realistic model of communication usage during disasters.…”

Section: Related Workmentioning

confidence: 99%

Realistic mobility simulation of urban mesh networks

2009

It is a truism that today's simulations of mobile wireless networks are not realistic. In realistic simulations of urban networks, the mobility of vehicles and pedestrians is greatly influenced by the environment (e.g., the location of buildings) as well as by interaction with other nodes. For example, on a congested street or sidewalk, nodes cannot travel at their desired speed. Furthermore, the location of streets, sidewalks, hallways, etc. restricts the position of nodes, and traffic lights impact the flow of nodes. And finally, people do not wander the simulated region at random, rather, their mobility depends on whether the person is at work, at lunch, etc. In this paper, realistic simulation of mobility for urban wireless networks is addressed. In contrast to most other mobility modeling efforts, most of the aspects of the presented mobility model and model parameters are derived from surveys from urban planning and traffic engineering research. The mobility model discussed here is part of the UDel Models, a suite of tools for realistic simulation of urban wireless networks. The UDel Models simulation tools are available online.

2010 IEEE Global Telecommunications Conference GLOBECOM 2010

“…There were many attempts at creating synthetic mobility patterns, ranging from methods based on connectivity graph [3], action profiles [2] to combining terrain and vehicle properties separately [4], to capturing group behavior [5], eventdriven [9], [10], [11], and finally to extraction of information from real world traces [6], [7].…”

Section: Previous Workmentioning

confidence: 99%

PCFG Based Synthetic Mobility Trace Generation

Geyik

Bulut

Szymanski

2010

This paper introduces a novel method of generating mobility traces based on Probabilistic Context Free Grammars (PCFGs). A PCFG is a generalization of a context free grammar in which each production rule is augmented with a probability with which this production is applied during sentence generation. A concise PCFG can be inferred from the given real world trace collected from the actual mobile node behaviors. The resulting grammar can be used to generate sequences of arbitrary length mimicking the mobile node behavior. This is important when new protocol designs for mobile networks are tested by simulation.In the paper, we describe the methods developed to construct such grammars from training data (mobility history). We also discuss how to generate the synthetic data with an already constructed grammar. We present the experimental results on two real data sets, measuring similarity of the actual traces with the synthetic ones. We compare our grammar based method to a 2-level Markov Model based trace generation method. The results demonstrate that the grammar based approach works as an excellent compression method for the actual data. On many metrics, the synthetic data generated from the PCFG match the training data much better than the one generated by the Markov Model.