IEEE CEDA 1st Internet-of-Things (IoT) Students Competition: “Smart and Pervasive Tracking and Monitoring in the IoT Scenario” Organization:   Organizers: José L. Ayala Pierre-Emmanuel Gaillardon UCM, Madrid, Spain EPFL, Lausanne, Switzerland [email protected] [email protected] CEDA Representative: David Atienza (VP of Conferences, EPFL) Scenario Smartphones with Wi-Fi enabled periodically transmit Wi-Fi messages, even when not associated to a network. In this contest, we propose to develop a software system for passively tracking unmodified smartphones, based on such Wi-Fi detections. This system will rely only on common, off-the-shelf access point hardware to both collect and deliver detections. To detect when Wi-Fi networks are available, these phones periodically scan the Wi-Fi band for access points. Every Wi-Fi transmission contains a unique device identifier (MAC address). Therefore, by taking advantage of the Wi-Fi monitoring capabilities of a laptop, it is possible to detect these transmissions in an area of interest, providing a coarse-grained location trace for each phone that passes through the area without having to embed any thirdparty software in the phones. In order to give some positioning capabilities to the passive tracking system, methods and algorithms that make use of the received signal strength indicator (RSSI) are expected. The experimentation scenario will consist on having the different teams assigned to known positions across a selected area of the conference center, but the specifically assigned locations will remain unknown for the participants in the contest before the day of the contest. Finally, in order to increase the accuracy of the passive tracking systems, teams will have to collaborate and integrate information gathered by other teams in different assigned positions. The data exchange will be done through Google spreadsheets. Setup and Evaluation The participants will be organized in teams of up to 3 students. The IoT contest organizers will provide in advance a map of the possibly assigned positions for the teams (with power plugs and WiFi internet access), something similar to: Exact positions will be disclosed on the day of the contest (Monday of the DATE week). Each team brings its own computer with the running tracking software and presents a poster with their solution (software engineering and algorithms). The teams are expected to use their own wireless interface to sniff the ‘hello packets’ of the smartphones. Programming languages and software infrastructures are free (all integrated C/C++ programs, or scripting environments). All this development will be accomplished by the participants before the contest on-site day. Data exchanged between the teams will be done using a Google spreadsheet (Google account to be given later on). The data format will be the following: Team position MAC address RSSI The evaluation of the proposed solutions will be performed with a set of controlled members (DATE Executive Committee members) who will transit across the area during the contest (MAC addresses to be known in advance). Specific tracking positions will be evaluated by the IoT coordinators. The data collected by the participants will be evaluated according to the achievement of the following functional levels: Level 1: Linked data (10pts) The application will be able to collect, annotate and store the data acquired from every wireless device in the area. Objectives Criteria of success Score Number of detected devices Exact number of DEC devices 5/5 Within 20% of the exact 3/5 Amount of information gathered per device Level 2: Localization (20pts) /5 In this level, the application also retrieves and stores information about real-time localization of the devices. Objectives Criteria of success Score Initial localization functionality Live console positions /5 Simple visualization app GUI visualization /5 Granularity in time of the tracking Response in less than 5s. /5 Quality of the algorithmic solution /5 Level 3: Notification system (10pts) In this level, the application will detect and understand “noisy” and outstanding situations in the monitored data, like connection problems, out-of-range users, conflicting data, etc. Objectives Criteria of success Score Number of detected outstanding situations /5 Capacity to infer knowledge from history /5 Level 4: Streaming publish/subscribe system (10pts) The collected data can be used to generate a set of data streams, where users can subscribe to a specific data or event (like proximity of a specific user, occupation of an area, etc.) to receive the required information. For that purpose, an end-user app is required as well. Objectives Criteria of success Score Capability to filter data streams Variety of criteria /5 Complex GUI interface Live web application /5 The team with the highest score will win the competition. The price will be subsequently given on Tuesday during the plenary session. From an organizational perspective, registration to the contest will close on January 15 th. The teams will have to send a description of the solution followed 2 weeks before the contest on-site day (4 pages report). On site, the contest will take place on a 4 hours slot, which comprises 2 hours of installation – assignment of positions, final debug, … and 2 hours of real test. The evaluation will be done during these 2 hours. Each team will also have to come with a poster describing their technical solution. The DATE attendees are encouraged to come and discuss with the students during this time. Technical resources Using Google spreadsheets as a database in cloud https://www.youtube.com/watch?v=F3nZM7YV2Y0 http://burnash.github.io/gspread/ https://github.com/google/google-api-cpp-client http://www.lbreda.com/grive/start Packet capture library (in C – exploited in wireshark) http://www.tcpdump.org/#documentation http://eecs.wsu.edu/~sshaikot/docs/lbpcap/libpcap-tutorial.pdf