lsinfo3/mim-stm
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|
Repository files navigation
This is a source traffic generator for mobile instant messaging traffic which generates for a given population size a contact network, i.e. a mapping of persons to groups via a max-flow problem. Then, those mappings are used to populate group chats with messages, media types, and the payload of the messages. solve_max_flow.py or solve_max_flow_n_runs.py are used for generating N repetitions for each population size in problem_sizes: - dat_generator.py generates for a given population size a pyomo input file as problem statement. - max_flow.py contains the pyomo max-flow problem definition with all constraints. - The resulting flow statistics are obtained in comparison_result.csv where the achieved and optimum flow are listed for each run for analysis. - From the results a mapping is saved into the mappings folder which is a json dictionairy. draw_messages.py: - reads the json mapping - generates a dataframe where each line is a chat - columns: - timestamps: message timestamps - num_messages: number of total messages - participation_dict: mapping of person ids to group participation - senders: sender_id of each message - media_types: media type of each message - message_sizes: message sizes in byte - message_payload: after compression with overhead in megabyte - sum_message_payloads_MB: sum of all message_payload of a chat - network_[scope]_MB: total network load when sending the message to the server and replicating it to each recipient within the scope