ZipWeave: Towards efficient and reliable measurement based mobile coverage maps

Abstract

The accuracy of measurement-driven mobile coverage maps depends on the quality, density and pattern of the signal strength observations. Thus, identifying an efficient measurement data collection methodology is essential, especially when considering the cost associated with the measurement collection approaches (e.g., drive tests, crowd approaches). We propose ZipWeave, a novel measurement data collection and fusion framework for building efficient and reliable measurement-based mobile coverage maps. ZipWeave incorporates a novel nonuniform sampling strategy to achieve reliable coverage maps with reduced sample size. Assuming prior knowledge of the propagation characteristics of the region of interest, we first examine the potential gains of this non-uniform sampling strategy in different cases via a measurement-based statistical analysis methodology; this involves irregular spatial tessellation of the region of interest into sub-regions with internally similar radio propagation characteristics and sampling based on these subregions. We then present a practical form of ZipWeave nonuniform sampling strategy that can be used even without any prior information. In all our evaluations, we show that the ZipWeave non-uniform sampling approach reduces the samples by half compared to the common systematic-random sampling, while maintaining similar accuracy. Moreover, we show that the other key feature of ZipWeave to combine high-quality controlled measurements (that present limited geographic footprint similar to drive tests) with crowdsourced measurements (that cover a wider footprint) leads to more reliable mobile coverage maps overall.

Improving Signal-Strength Aggregation for Mobile Crowdsourcing Scenarios

“…There are some works that developed high-resolution coverage maps from Mobile Crowdsourcing signal-strength measurements. These maps were created by plotting each empirical sample on the map [ 23 , 24 ] or by interpolating the signal strength in several uniformly distributed points inside the area of interest using linear interpolation [ 25 ], using variations of Kriging method [ 26 , 27 ] or by using Gaussian processes that consider a prior knowledge about theoretical path loss models [ 28 ]. These coverage maps are useful for tasks that require highly detailed maps, but when analyzing signal coverage in greater areas, the effectiveness of their fine-grained visualizations will decrease as the resolution of the maps decreases.…”

“…Indeed, many of these papers manipulated dBm values without mentioning the correspondence between dBm and watt values [ 13 , 15 , 20 , 21 , 27 , 36 , 37 , 38 , 39 , 40 , 41 ], and moreover, some of them manipulated signal-strength values without reporting the unit of measurement employed [ 12 , 14 , 29 , 42 , 43 ]. Many of the papers that followed these wrong methodologies got log-scaled signal-strength measurements directly from mobile operating systems (Android or iOS) [ 1 , 8 , 11 , 12 , 13 , 15 , 20 , 23 , 26 , 27 , 36 , 37 , 38 , 40 ]. Therefore, it is plausible that they just used and manipulated the data returned by the systems without a thorough analysis about the unit of the collected signal-strength values.…”

“…This is performed by calculating different measures of prediction accuracy as MAE (Mean Absolute Error), MAPE (Mean Absolute Percentage Error), or MASE (Mean Absolute Scaled Error). Nevertheless, many of these works failed at estimating their prediction accuracy by using a logarithmic scale for prediction errors, and summarizing them by applying mean-based aggregation as MAPE [ 26 , 27 ], MSE (Mean Squared Error) [ 28 ], or RMSE (Root Mean Squared Error) [ 45 ]. Then, they added a source of error at applying mean functions to log-scaled values, as mentioned in Section 3.1 .…”

Results from Running an Experiment as a Service Platform for Mobile Networks

“…Also, reliance on users can provide high coverage, but at the cost of repeatability regarding location, route or equipment. However, in combination with a platform like MONROE, they could be used in a more systematic and controllable way [8].…”

“…Tstat [9] is a powerful passive monitoring tool that rebuilds TCP flows reporting more than 100 flow descriptors (e.g., client and server IP and port, RTT, number of retransmissions) and more than a thousand packet level metrics. 7 Experimenters can use Graphite 8 to easily navigate through the logs and store a dashboard showing relevant data within an adjustable time window.…”