Dataset for "Adaptive Content Seeding for Information-Centric Networking under High Topology Dynamics"

Dataset to accompany the manuscript "Adaptive Content Seeding for Information-Centric Networking under High Topology Dynamics".
(Version 2)

Instructions:
The files contained in this dataset measure the following metrics:

Cache miss ratio: this metric, computed by every requester, is calculated as the ratio between the number of non-satisfied content requests (i.e., interests sent via \ac{D2D} links) and the total number of interest messages sent. The metric is recorded in the following files: CacheMissRatio_Low.csv (low density), CacheMissRatio_Medium.csv (medium density), CacheMissRatio_High.csv (high density).

Content hop count: this metric is computed by every requester and measures the number of hops to reach the content via \ac{D2D} communication links. The metric is recorded in the following files: CacheHitHops_Low.csv (low density), CacheHitHops_Medium.csv (medium density), CacheHitHops_High.csv (high density).

Sent control messages: this metric is computed by every vehicle by counting the total number of sent interest and acknowledgment messages. The metric is recorded in the following files: SentCtrlMsgs_Low.csv (low density), SentCtrlMsgs_Medium.csv (medium density), SentCtrlMsgs_High.csv (high density).

Content downloads: this metric is computed by the centralized controller and counts the number of content downloads from the backend service via RAN for all vehicles during one simulation run. The metric is recorded in the following files: ContentDownloads_Low.csv (low density), ContentDownloads_Medium.csv (medium density), ContentDownloads_High.csv (high density).

Channel busy ratio: this metric is computed by every vehicle and measures the ratio of time the channel is sensed busy over the total active simulation time. The metric is recorded in the following files: CBR_Low.csv (low density), CBR_Medium.csv (medium density), CBR_High.csv (high density).

Every file in the dataset has the following columns: <"Strategy","Clustering","reqProb","Value","sd","se","ci">

- "Strategy" contains the employed seeding algorithm
- "Clustering" contains the employed community detection algorithm
- "reqProb" contains the request probability p
- "Value" contains the mean value of the actual metric that is being measured, which is defined in the file name
- "sd","se","ci" are the standard deviation, standard error of the mean, and confidence interval (95%)

Changes with respect to the previous version:

- We repeated every simulation run using 100 different seeds, instead of 50 as in the previous version of the dataset
- We deleted the "Number of clusters" metric, as it is not being used in the manuscript