MNMA Dataset

This document is a description of a dataset which can be used for mobile network management automation (MNMA) related research scenarios. This dataset has been generated using a Nokia internal simulator. The simulation scenario is a Manhattan grid model which contains 1000 users distributed across an area of 4 sq kilometers and 5 cells. There is a central cell which is overlapped by four other cells in four sides. The users were all of different types like pedestrians, bike riders, cars, etc. and the user mobility was random. The data was collected at the cell level and it contains: (i) downlink throughput in bits/s, (ii) radio link failures, (iii) cell load, (iv) handover attempts and (v) successful handovers. The data was collected every 10 seconds for a duration of 6 hours at each cell, so the simulation ran for $3600$ time steps. KPIs of each cell were recorded separately and they were dumped into text files in JSON format. To observe handover related performances and KPIs, it is recommended to use the data measured in the centrall cell (cell id: 4).

For more information and quantitative analysis of the dataset, please refer to the following paper: 

A. Banerjee, S. S. Mwanje and G. Carle, "Toward Control and Coordination in Cognitive Autonomous Networks," in IEEE Transactions on Network and Service Management, vol. 19, no. 1, pp. 49-60, March 2022, doi: 10.1109/TNSM.2021.3116308.