mobIPLity

Welcome to the mobIPLity trace file generator. mobIPLity produces bonnmotion traces from the eduroam records collected by the Instituto Politécnico de Lisboa (IPL). Its goal is to provide to the mobile computing research community realistic mobility scenarios.

MobIPLity was first described in a paper published in Mobiquitous 2014 [1]. Previous characterisations and analysis of the data set can be found in [2], [3] and [4].

Data set characterisation

Mobility scenarios are created from the complete set of more than 31 million access records captured by the 200+ access points of the eduroam node at the IPL between Janury 1st, 2005 and December 31st, 2012. Overall, the date set considers more than 48000 distinct devices, carried by over 30000 users.

IPL is spread over 7 distinct campus in the Lisbon metropolitan area (see maps). You can select mobility scenarios from any of the campus or from the entire IPL, knowing in advance that the distinct scientific areas covered by each campus (health sciences, engineering, etc.) result in distinct usage and mobility patterns.

You can also arrange mobility scenarios by one of three categories of devices:

  • Small devices (i.e. smartphones, PDAs, tablets),
  • Laptops
  • Any device

Devices where arranged according to the OS identification exhibited in DHCP records. The "Any devices" option combines small devices, laptops and the remaining devices which were not possible to associate to any of the remaining cathegories.

Instructions

The generation of scenarios requires the specifiation of the following parameters:

Start/End date

Scenario must not start/end before/after the dates indicated. The effective start date is the first moment after start date that satisfies the remaining constraints. As a measure to protect user privacy, the effective start date is never disclosed.

Min points

Suggested: 3. Routes must have passed by at least this number of access points

Min devices

Pick a scenario from a moment where at least "min devices" where simultaneously observed.

Device type

Puts in the scenario exclusively devices that:
Device TypeDefinition
SMD (Small Mobile Devices)Those whose DHCP messages claimed to be running Android, Windows Mobile or iOS operating system
LaptopsThose whose DHCP messages claimed to be running Linux, Windows or MAC OS Operating System
AllAny device, thus including Laptops and SMDs but also those whose DHCP messages claimed some unidentified OS

Number of Axes

Suggested: 2D. Includes altitude in the location coordinates.

Site

Select one of the campus:
Campus (field of knowledge)Scenario Area (km2,approx.)
ISEL (Engineering)0.063
ESEL (Education)0.013
ESTC (Theater and Film)0.009
ESML (Music)0.008
ESTESl (Health Technology)0.007
ESCS (Communication and Media)0.006
ESD (Dance)0.003
SC (Presidency)0.002
RMB (Students Dorm)0.002
SAS (Social Services)0.001
ISCAL (Accounting and Administration)0.001
All (considering Lisbon as a campus)40

Duration

Duration of the scenario to extract. (Suggested: 120min)

Warmup/Coolof

Devices must have one or more records in the "warmup" seconds before (resp. after) the beginning (resp. the end) of the scenario (Suggested: 120min)

Enhanced trace

If yes, scenario output will follow disaster area (DA) syntax. Which will include options to mark a device as being network disconnected and results in more realistic simulations, reflecting operating systems mechanisms for extending battery life. If you choose "no", devices that disconnect during the scenario duration will not be included in the scenario. Note that network simulators may react differently to this feature. (Suggested: no)

AP range

Estimated coverage range for Access Points. mobIPLity defines devices routes considering the location and coverage of the access points. This is a configuration parameter that can be used to tune the device locations. See [1] for an explanation. (Suggested: 50m)

Min speed

When disconnected from the network, a device can be considered to be disconnected or moving. The choice will be made automatically by mobIPLity and will depend of the speed of the device, for moving between the location of one AP and the next one. If the observed speed is below this threshold, the device will be considered to be disconnected. (Suggested: 0.5 m/s)

Reconnect time

If a device disconnects and later reconnects to the same AP, it can be assumed that either the trace for the device was ended (e.g. user went home and came back in the next day) or that the device is simply saving battery. The former case will be assumed if the time between the disconnection and the reconnection is above this threshold value. Used in conjuntion with "Min speed", a device has to be faster than "Min speed" OR take less than "Reconnect time". (Suggested: 120s)

How It Works

The specified parameters are passed to the mobIPLity scenario generator which will search for the first date satisfying all the constraints given. A file with the scenario will be prepared offline and sent by e-mail to the requester as soon as possible. To prevent abuses or any attempt to violate user's privacy, all requests are moderated by a human operator.

Scenario Generation

Requester identification:

Traceset options (see example here):

References

[1] The impact of device type agnostic mobility models on ubiquitous applications.
Nuno Cruz and Hugo Miranda. Proceedings of Mobiquitous 2014
[2] The evolution of user mobility on the eduroam network
Nuno Cruz, Hugo Miranda and Pedro Ribeiro. Proceedings of the 2014 IEEE International Conference on Pervasive Computing and Communications Workshops (PERCOM Workshops). Budapest, Hungary. March 24, 2014. IEEE
[3] O impacto dos smartphones nos modelos de mobilidade tradicionais
Nuno Cruz, Hugo Miranda, and Pedro Ribeiro. INForum 2014 - Atas do 6º Simpósio de Informática. pp. 195-210, Porto, Portugal, September 4-5, 2014.
[4] A mobilidade dos utilizadores da eduroam ao longo dos anos
Nuno Cruz, Hugo Miranda, and Pedro Ribeiro. Atas do 5o Simpósio de Informática (INForum 2013). João Cachopo and Beatriz Sousa Santos (eds.) pp. 189-200. Universidade de Évora, Portugal. Sept. 5-6, 2013.

Contact

Drop us a line if you like and/or have any doubts about our work by clicking here.