Virtual Delivery Trees Evaluation Results

<p>The artifacts represent evaluation results of real world networks having more than 40 nodes from <a href="http://www.topology-zoo.org/">Network Topology Zoo</a>. The applied topologies are listed in following table, sorted in descending order by diameter (d) and nodal degree fluctuation (&sigma;^2)).</p> <table> <thead> <tr> <th scope="col">Topology</th> <th scope="col">|V|</th> <th scope="col">|E|</th> <th scope="col">&lt;k&gt;</th> <th scope="col">&sigma;^2</th> <th scope="col">d</th> </tr> </thead> <tbody> <tr> <td>Chinanet</td> <td>42</td> <td>66</td> <td>1.5</td> <td>10.52</td> <td>4</td> </tr> <tr> <td>Litnet</td> <td>43</td> <td>43</td> <td>0.98</td> <td>5.04</td> <td>4</td> </tr> <tr> <td>Cernet</td> <td>41</td> <td>58</td> <td>1.32</td> <td>5.6 |</td> <td>5 |</td> </tr> <tr> <td>Ntt</td> <td>32</td> <td>65</td> <td>1.48</td> <td>7.07</td> <td>6</td> </tr> <tr> <td>Cesnet200706</td> <td>44</td> <td>51</td> <td>1.16</td> <td>6.27</td> <td>6</td> </tr> <tr> <td>Carnet</td> <td>44</td> <td>43</td> <td>0.98</td> <td>5.48</td> <td>6</td> </tr> <tr> <td>Dfn</td> <td>50</td> <td>78</td> <td>1.77</td> <td>5.31</td> <td>6</td> </tr> <tr> <td>Telcove</td> <td>71</td> <td>70</td> <td>1.59</td> <td>9.13</td> <td>7</td> </tr> <tr> <td>Forthnet</td> <td>62</td> <td>62</td> <td>1.41</td> <td>7.72</td> <td>7</td> </tr> <tr> <td>Bellsouth</td> <td>51</td> <td>66</td> <td>1.5 |</td> <td>7.55 |</td> <td>7 |</td> </tr> <tr> <td>Garr200902</td> <td>54</td> <td>68</td> <td>1.55</td> <td>5.13</td> <td>7</td> </tr> <tr> <td>Arnes</td> <td>41</td> <td>57</td> <td>1.3 |</td> <td>4.53 |</td> <td>7 |</td> </tr> <tr> <td>BeyondTheNetwork</td> <td>53</td> <td>65</td> <td>1.48</td> <td>3.98</td> <td>7</td> </tr> <tr> <td>Uunet</td> <td>49</td> <td>84</td> <td>1.91</td> <td>7.38</td> <td>8</td> </tr> <tr> <td>Tw</td> <td>71</td> <td>115</td> <td>2.61</td> <td>| 5.58</td> <td>| 8</td> </tr> <tr> <td>Uninett</td> <td>71</td> <td>97</td> <td>2.2</td> <td>3.12</td> <td>9</td> </tr> <tr> <td>Renater2010</td> <td>43</td> <td>56</td> <td>1.27</td> <td>3.08</td> <td>9</td> </tr> <tr> <td>Surfnet</td> <td>50</td> <td>68</td> <td>1.55</td> <td>3.36</td> <td>11</td> </tr> <tr> <td>Iris</td> <td>51</td> <td>64</td> <td>1.45</td> <td>2.16</td> <td>11</td> </tr> <tr> <td>Palmetto</td> <td>45</td> <td>64</td> <td>1.45</td> <td>2.57</td> <td>12</td> </tr> <tr> <td>BtLatinAmerica</td> <td>45</td> <td>50</td> <td>1.14</td> <td>1.87</td> <td>12</td> </tr> <tr> <td>Bellcanada</td> <td>48</td> <td>64</td> <td>1.45</td> <td>2.59</td> <td>13</td> </tr> <tr> <td>Sanet</td> <td>43</td> <td>45</td> <td>1.02</td> <td>1.66</td> <td>13</td> </tr> <tr> <td>LambdaNet</td> <td>42</td> <td>46</td> <td>1.05</td> <td>1.57</td> <td>13</td> </tr> <tr> <td>HiberniaGlobal</td> <td>55</td> <td>81</td> <td>1.84</td> <td>2.72</td> <td>16</td> </tr> <tr> <td>Ntelos</td> <td>47</td> <td>58</td> <td>1.32</td> <td>1.92</td> <td>17</td> </tr> <tr> <td>RedBestel</td> <td>84</td> <td>93</td> <td>2.11</td> <td>0.85</td> <td>28</td> </tr> <tr> <td>VtlWavenet2008</td> <td>88</td> <td>92</td> <td>2.09</td> <td>0.11</td> <td>31</td> </tr> </tbody> </table> <p>The evaluation results consist of three major parts:</p> <ol> <li><em>Raw Data</em>: Configuration and results of all simulation experiments as CSV files.</li> <li><em>Strategy Results</em>: Visualization of the test results for each topology.</li> <li><em>Best Strategies</em>: Highlighting of the best strategies across all topologies.</li> </ol> <p>Therein, the &quot;Raw Data<em>&quot;</em> comprise the configuration of or simulation experiments and the simulation results. Each line stands for a single simulation run.</p> <p><em>&quot;</em>Strategy Results<em>&quot; </em>and &quot;Best Strategies&quot; accompany the results presented in the paper. Result plots in the paper are excerpts from the plots in this repository. See below for further details.</p> <p><strong>Raw Data</strong></p> <p>Both, the configuration of a run and its results correspond to one line within a CSV file in subfolder <code>./raw</code>. Each file comprises the results of a replication.<br> &nbsp;</p> <pre><code>raw ├── results_0.csv ├── results_1.csv ├── ... └── results_9.csv</code></pre> <p>The raw data of a CSV file is structured as follows.</p> <table> <thead> <tr> <th scope="col">Column</th> <th scope="col">Description</th> </tr> </thead> <tbody> <tr> <td>topo</td> <td>Topology name.</td> </tr> <tr> <td>peers</td> <td>Number of nodes.</td> </tr> <tr> <td>edges</td> <td>Number of links.</td> </tr> <tr> <td>p_publishers</td> <td>Proportion of nodes acting as publisher (15% - 45%).</td> </tr> <tr> <td>p_subscriber</td> <td>Proportion of nodes acting as subscriber (15% - 45%).</td> </tr> <tr> <td>n_rules</td> <td>Number of allowed rules per switch.</td> </tr> <tr> <td>distances</td> <td>Flag for consideration of geographical distances (currently not used).</td> </tr> <tr> <td>strategy</td> <td>Applied virtual tree strategy.</td> </tr> <tr> <td>distribution</td> <td>Distribution method for client (uniform, distant, nearby)</td> </tr> <tr> <td>n_cluster</td> <td>Number of simulated clusters within the topology.</td> </tr> <tr> <td>p_change</td> <td>Churn rate of clients (0% - 100%).</td> </tr> <tr> <td>pub_change</td> <td>Flag for publisher migration (currently not used).</td> </tr> <tr> <td>tree_count</td> <td>Number of virtual trees installed in the network.</td> </tr> <tr> <td>selected_subscribers</td> <td>Avg. number of subscribers addressed by a publisher</td> </tr> <tr> <td>init_cost</td> <td>Avg. number of entries of a non-optimized distribution tree (per notification)</td> </tr> <tr> <td>trees</td> <td>Avg. proportion of tree entries per notification.</td> </tr> <tr> <td>stops</td> <td>Avg. proportion of stop entries per notification.</td> </tr> <tr> <td>hops</td> <td>Avg. proportion of hop entries per notification.</td> </tr> <tr> <td>final_cost</td> <td>Aggregated proportions (trees + stops + hops).</td> </tr> <tr> <td>datetime</td> <td>Timestamp of the simulation run.</td> </tr> </tbody> </table> <p><strong>Result Charts</strong></p> <p>The simulation results are visualized in <code>plots.md</code> or <code>plots.html</code>, ordered according above topology table.</p> <p>Each topology accompanys following:<br> - Topology figures with the computed <em>Clusters</em> therein.<br> - Line charts outlining the behavior of the strategies over changing <em>Number of Flow Rules</em>.<br> - Bar charts outlining the strategies&#39; performance for different <em>Migration Scenarios</em>.</p> <p>Details of the figures and diagrams are described next.</p> <ul> <li>Clusters:<br> Visualization of exemplary groups within the topology, computed by `clusters` and `partitions` strategy. The clusters strategy assigns 60% of a network&#39;s nodes to cluster groups; the partition strategy, in contrast, assigns all nodes to groups. Both strategies are described in Sec. III.<br> &nbsp;</li> <li>Number of Flow Rules:<br> Results for varying number of rules (from 5 to 40) per switch, as described in Sec. IV. The charts are organized in a 3 x 3 matrix. A row of the matrix corresponds to different proportions of subscribers per publisher (15%, 30%, and 45%); a column corresponds to different distributions of clients (uniform, nearby and distant).<br> &nbsp;</li> <li>Migration Scenarios:<br> Results for different migration scenarios with a fixed number of rules (40 rules per switch), as described in Sec. V. Each bar group stands for a strategy and reflects the results of different migration rates (0%, 30%, 50%, 70%, 100%).</li> </ul> <p><strong>Best Strategies</strong></p> <p>Scatter plots in subfolder <code>./fluctuation</code> visualize the most efficient strategies for different migration scenarios by considering different proportions of subscribers per publisher (15%, 30%, and 45%). The plots show the results for a fixed number of subscribers (30% per publisher) and a churn rate of 100%. The strategies therein require the fewest labels in the header stack to encode a notification distribution tree, represented by the strategy&#39;s dot size</p>